report on the deliberation results - 医薬品医療機器総合機構 · 2020-07-20 · market...

This English translation of this Japanese review report is intended to serve as reference material made available for the convenience of users. In the event of any inconsistency between the Japanese original and this English translation, the Japanese original shall take precedence. PMDA will not be responsible for any consequence resulting from the use of this reference English translation.

Report on the Deliberation Results

November 30, 2018

Pharmaceutical Evaluation Division, Pharmaceutical Safety and Environmental Health Bureau

Ministry of Health, Labour and Welfare

Brand Name Keytruda Injection 20 mg

Keytruda Injection 100 mg

Non-proprietary Name Pembrolizumab (genetical recombination) (JAN*)

Applicant MSD K.K.

Date of Application March 30, 2018

Results of Deliberation

In its meeting held on November 29, 2018, the Second Committee on New Drugs concluded that the partial

change application for the product may be approved and that this result should be presented to the

Pharmaceutical Affairs Department of the Pharmaceutical Affairs and Food Sanitation Council.

The re-examination period for the indication of “advanced or recurrent microsatellite instability-high (MSI-

High) solid tumors that have progressed after cancer chemotherapy” is 4 years. The re-examination period for

the indication of “malignant melanoma” is the remainder of the ongoing re-examination period for the initial

approval (until September 27, 2026).

Conditions of Approval

[Advanced or recurrent MSI-High solid tumors that have progressed after cancer chemotherapy (limit the use

to patients who are refractory or intolerant to standard treatments)]

1. The applicant is required to develop and appropriately implement a risk management plan.

2. The applicant is required to provide healthcare professionals with the results of 2 ongoing phase II studies

in patients with advanced or recurrent MSI-High solid tumors that have progressed after cancer

chemotherapy, promptly after the studies are completed.

3. Since only limited data are available as to the efficacy of the product in the treatment of MSI-High solid

tumors (except colorectal cancer), the applicant is required to conduct a drug use-results survey after the

market launch to collect information on the characteristics of patients treated with the product, to promptly

collect data on the efficacy and safety of the product, and to take necessary actions to ensure the proper

use of the product.

[Malignant melanoma]

The applicant is required to develop and appropriately implement a risk management plan.

*Japanese Accepted Name (modified INN)


Keytruda Injection (MSI-High Solid Tumors, etc.)_MSD K.K._review report

Review Report

November 19, 2018

Pharmaceuticals and Medical Devices Agency

The following are the results of the review of the following pharmaceutical product submitted for marketing

approval conducted by the Pharmaceuticals and Medical Devices Agency (PMDA).

Brand Name Keytruda Injection 20 mg Keytruda Injection 100 mg Non-proprietary Name Pembrolizumab (genetical recombination) Applicant MSD K.K.

Date of Application March 30, 2018, May 10, 2018, and August 9, 20181)

Dosage Form/Strength Injection: Each 0.8 mL vial contains 20 mg of pembrolizumab (genetical

recombination).

Injection: Each 4 mL vial contains 100 mg of pembrolizumab (genetical

recombination).

Application Classification Prescription drug, (4) Drug with new indications, (6) Drug with a new dosage

Items Warranting Special Mention

Orphan drug (Orphan Drug Designation No. 350 of 2014 [26 yaku]; PFSB/ELD

Notification No. 0917-6 dated September 17, 2014, by the Evaluation and

Licensing Division, Pharmaceutical and Food Safety Bureau, Ministry of Health,

Labour and Welfare)

Drug for the conditional early approval system (PSEHB/PED Notification No.

0622-2 dated June 22, 2018, by the Pharmaceutical Evaluation Division,

Pharmaceutical Safety and Environmental Health Bureau, Ministry of Health,

Labour and Welfare)

Priority review (PSEHB/PED Notification No. 0703-3 dated July 3, 2018, by the

Pharmaceutical Evaluation Division, Pharmaceutical Safety and Environmental

Health Bureau, Ministry of Health, Labour and Welfare)

Reviewing Office Office of New Drug V

Results of Review

On the basis of the data submitted, PMDA has concluded that the product has efficacy for the following

indications, and that the product has acceptable safety in view of its benefits (see Attachment). PMDA has also

1) Partial change approval applications for (a) a new indication for the treatment of MSI-High solid tumors and a modified dosage and

administration for the treatment of unresectable malignant melanoma, (b) a new indication and a new dosage and administration for the adjuvant treatment of malignant melanoma, and (c) a modified indication for the treatment of NSCLC were filed on (a) March 30, 2018, (b) May 10, 2018, and (c) August 9, 2018, respectively.


Keytruda Injection (MSI-High Solid Tumors, etc.)_MSD K.K._Review Report

concluded that the change in the dose of pembrolizumab (genetical recombination) from a body weight-based

dose to a fixed-based dose for the treatment of unresectable malignant melanoma is acceptable.

• Pembrolizumab (genetical recombination) monotherapy for the adjuvant treatment of malignant melanoma

• Pembrolizumab (genetical recombination) in combination with platinum-based chemotherapy for the

treatment of chemotherapy-naïve patients with unresectable, advanced or recurrent non-small cell lung

cancer

• Pembrolizumab (genetical recombination) monotherapy for the treatment of chemotherapy-naïve patients

with unresectable, advanced or recurrent PD-L1 positive (Tumor Proportion Score [TPS] 2) ≥1%) non-small

cell lung cancer

• Pembrolizumab (genetical recombination) monotherapy for the treatment of chemotherapy-treated patients

with advanced or recurrent microsatellite instability-high (MSI-High) solid tumors, who have no other

standard treatment options

As a result of its review, PMDA has concluded that the product may be approved for the indication and dosage

and administration shown below, with the following conditions. The efficacy and safety of the product in the

treatment of MSI-High solid tumors (except colorectal cancer) should be further investigated through post-

marketing surveillance.

Indications

Unresectable Malignant melanoma

Unresectable, advanced or recurrent PD-L1 positive non-small cell lung cancer

Relapsed or refractory classical Hodgkin lymphoma

Radically unresectable urothelial carcinoma that has progressed after cancer chemotherapy

Advanced or recurrent microsatellite instability-high (MSI-High) solid tumors that have progressed after

cancer chemotherapy (limit the use to patients who are refractory or intolerant to standard treatments)

(Underline denotes additions. Strikethrough denotes deletions.)

Dosage and Administration

[Unresectable Malignant melanoma]

The usual adult dosage is 200 mg 2 mg/kg body weight of pembrolizumab (genetical recombination) infused

intravenously over 30 minutes every 3 weeks. For the adjuvant treatment of malignant melanoma, the

maximum duration of treatment is 12 months.

[Unresectable, advanced or recurrent PD-L1 positive non-small cell lung cancer; relapsed or refractory

classical Hodgkin lymphoma; radically unresectable urothelial carcinoma that has progressed after cancer

chemotherapy; or advanced or recurrent MSI-High solid tumors that have progressed after cancer

chemotherapy (limit the use to patients who are refractory or intolerant to standard treatments)]

2) Percentage of cells expressing PD-L1 in the tumor tissue


Keytruda Injection (MSI-High Solid Tumors, etc.)_MSD K.K._Review Report

The usual adult dosage is 200 mg of pembrolizumab (genetical recombination) infused intravenously over 30

minutes every 3 weeks.

(Underlines denote additions. Strikethrough denotes deletions.)


[Malignant melanoma, or unresectable, advanced or recurrent non-small cell lung cancer]


[Advanced or recurrent microsatellite instability-high (MSI-High) solid tumors that have progressed after

cancer chemotherapy (limit the use to patients who are refractory or intolerant to standard treatments)]


2. The applicant is required to provide healthcare professionals with the results of 2 ongoing phase II studies




tumors (except colorectal cancer), the applicant is required to conduct a drug use-results survey after the

market launch to collect information on the characteristics of patients treated with the product, to promptly

collect data on the efficacy and safety of the product, and to take necessary actions to ensure the proper use

of the product.

1


Attachment

Review Report (1)

October 12, 2018

The following is an outline of the data submitted by the applicant and content of the review conducted by the

Pharmaceuticals and Medical Devices Agency (PMDA).

Product Submitted for Approval Brand Name Keytruda Injection 20 mg Keytruda Injection 100 mg Non-proprietary Name Pembrolizumab (genetical recombination) Applicant MSD K.K.

Date of Application March 30, 2018, May 10, 2018, and August 9, 20181)

Dosage Form/Strength Injection: Each 0.8 mL vial contains 20 mg of pembrolizumab (genetical

recombination).

Injection: Each 4 mL vial contains 100 mg of pembrolizumab (genetical

recombination).

Proposed Indications Unresectable Malignant melanoma



Radically unresectable urothelial carcinoma that has progressed after cancer

chemotherapy

Locally advanced or metastatic microsatellite instability-high (MSI-High) cancers

(Underline denotes additions. Strikethrough denotes deletions.)

Proposed Dosage and Administration

[Unresectable malignant melanoma]

The usual adult dosage is 2 mg/kg body weight of pembrolizumab (genetical

recombination) infused intravenously over 30 minutes every 3 weeks.

[Unresectable, advanced or recurrent PD-L1 positive non-small cell lung cancer;

relapsed or refractory classical Hodgkin lymphoma; or radically unresectable

urothelial carcinoma that has progressed after cancer chemotherapy]

The usual adult dosage is 200 mg of pembrolizumab (genetical recombination)

infused intravenously over 30 minutes every 3 weeks.

(Strikethrough denotes deletions.)

1) Partial change applications for (a) a new indication for the treatment of MSI-High solid tumors and a modified dosage and

administration for the treatment of unresectable malignant melanoma, (b) a new indication and a dosage and administration for the adjuvant treatment of malignant melanoma, and (c) a modified indication for the treatment of NSCLC were filed on (a) March 30, 2018, (b) May 10, 2018, and (c) August 9, 2018, respectively.

2


Table of Contents

1. Origin or History of Discovery, Use in Foreign Countries, and Other Information ................................... 3

2. Data Relating to Quality and Outline of the Review Conducted by PMDA .............................................. 6

3. Non-clinical Pharmacology and Outline of the Review Conducted by PMDA .......................................... 6

4. Non-clinical Pharmacokinetics and Outline of the Review Conducted by PMDA .................................... 7

5. Toxicity and Outline of the Review Conducted by PMDA ........................................................................ 7

6. Summary of Biopharmaceutic Studies and Associated Analytical Methods, Clinical Pharmacology, and

Outline of the Review Conducted by PMDA ............................................................................................. 7

7. Clinical Efficacy and Safety and Outline of the Review Conducted by PMDA ........................................ 7

8. Results of Compliance Assessment Concerning the New Drug Application Data and Conclusion Reached

by PMDA .................................................................................................................................................. 90

9. Overall Evaluation during Preparation of the Review Report (1) ............................................................ 90

List of Abbreviations

See Appendix

3


1. Origin or History of Discovery, Use in Foreign Countries, and Other Information

1.1 Outline of the proposed product

CD279 (programmed cell death-1 [PD-1]) is a receptor belonging to the CD28 superfamily (a group of

molecules that provide costimulatory signals involved in the control of T-cell activation) and is expressed on

activated lymphocytes (including T cells, B cells, and natural killer T cells). PD-1 in vivo is thought to bind to

PD-Ls (CD274 [PD-L1] and CD273 [PD-L2]) expressed on antigen-presenting cells to suppress the immune

response (Immunol Rev. 2010;236:219-42). PD-L1 and PD-L2 are expressed on a wide range of tumor tissues

(Nat Rev Immunol. 2008;8:467-77), suggesting that the PD-1/PD-L pathway is one of the mechanisms used by

tumor cells to avoid attacks by antigen-specific T cells.

Pembrolizumab (genetical recombination) (hereinafter referred to as “pembrolizumab”), a humanized

immunoglobulin (Ig) G4 monoclonal antibody against human PD-1, was discovered by the UK Medical

Research Council. Pembrolizumab binds to the extracellular domain of PD-1 (PD ligand binding site) and

blocks PD-1 binding to its ligands, PD-L1 and PD-L2. Pembrolizumab thereby enhances the activation of

cancer antigen-specific T cells and cytotoxic activation against cancer cells, resulting in the suppression of

tumor growth.

In Japan, pembrolizumab was approved for the treatment of “unresectable malignant melanoma” in September

2016, for the treatment of “unresectable, advanced or recurrent PD-L1 positive non-small cell lung cancer” in

December 2016, for the treatment of “relapsed or refractory classical Hodgkin's lymphoma” in November 2017,

and for the treatment of “radically unresectable urothelial carcinoma that has progressed after cancer

chemotherapy” in December 2017.

Recently, the applicant almost simultaneously submitted the following partial change approval applications (a)

to (c) on different days in a short period. This review report summarizes the reviews of these applications. In

this report, “MSI-High patients” are expressed as (a) “MSI-High (polymerase chain reaction [PCR]) patients,”

when the term refers to patients in whom a failure of the DNA mismatch repair system was detected by PCR

(narrow term), or (b) “MSI-High patients,” without specifying the assay method, when the term refers to

patients in whom a deficiency of DNA mismatch repair (MMR) was detected by either IHC or PCR (broad

term) [see Section 7.3.R.2.1 and other relevant sections].

(a) March 30, 2018

(i) A new indication for the treatment of microsatellite instability-high (MSI-High) solid tumors

(ii) A modified dosage and administration for the treatment of “unresectable malignant melanoma”

(b) May 10, 2018

A new indication and a new dosage and administration for the adjuvant treatment of malignant melanoma

(c) August 9, 2018

A modified indication for the treatment of “unresectable, advanced or recurrent PD-L1 positive non-small

cell lung cancer”

4


1.2 Development history, the approval status in foreign countries, and other information

1.2.1 Malignant melanoma: (a) (ii) and (b) in Section 1.1. See Section 7.1 for outline of the review

conducted by PMDA.

Pembrolizumab has been approved as a body weight-based dose (2 mg/kg every 3 weeks) therapy for

“unresectable malignant melanoma,” and as a fixed dose (200 mg every 3 weeks) therapy for the other approved

indications. The applicant has filed a partial approval application to change the dosage and administration for

the treatment of “unresectable malignant melanoma” from the approved body weight-based dose (2 mg/kg

every 3 weeks) to a fixed dose (200 mg every 3 weeks), primarily based on the results of a PPK analysis.

Furthermore, the applicant started a global phase III study of pembrolizumab monotherapy in patients with

resected malignant melanoma (Study 054) in ** 20**. In Japan, patient enrollment in the study was started in

** 20**. The applicant has filed a partial approval application for a new indication of the adjuvant treatment

for malignant melanoma, based on the pivotal data from Study 054.

Pembrolizumab was designated as an orphan drug in September 2014 with the intended indication of

“malignant melanoma” (Drug Designation No. 350 of 2014 [26 yaku]).

As of August 2018, the approval status of pembrolizumab outside Japan is as follows:

• Approval applications for the fixed-dose (200 mg every 3 weeks) treatment of unresectable malignant

melanoma were filed primarily based on the results of a PPK analysis in the US in August 2016 and in the

EU in April 2018, and approval was granted in May 2017 and August 2018, respectively. Pembrolizumab

has been approved for the fixed-dose treatment of unresectable malignant melanoma in 56 countries and

regions.

• Approval applications for the adjuvant treatment of malignant melanoma were filed in the US and EU in

April 2018, based on the pivotal data from Study 054. The applications are currently under review.

Pembrolizumab has not been approved for the adjuvant treatment of malignant melanoma in any country

or region.

1.2.2 NSCLC: (c) in Section 1.1. See Section 7.2 for outline of the review conducted by PMDA.

Pembrolizumab has been approved for the treatment of “unresectable, advanced or recurrent PD-L1 positive

non-small cell lung cancer,” and has also been approved as a monotherapy for chemotherapy-naïve patients

with PD-L1 positive (TPS ≥50%) NSCLC.

The applicant started the following 3 global phase III studies in patients with chemotherapy-naïve patients with

unresectable, advanced or recurrent NSCLC in (a) February 2016, (b) August 2016, and (c) ** 20**,

respectively, and in Japan, patient enrollment was started in (a) ** 20**, (b) ** 20**, and (c) ** 20**,

respectively.

(a) A clinical study of pembrolizumab in combination with platinum-based chemotherapy in patients with

nonsquamous NSCLC (NSQ-NSCLC) (Study 189)

5


(b) A clinical study of pembrolizumab in combination with platinum-based chemotherapy in patients with

squamous NSCLC (SQ-NSCLC) (Study 407)

(c) A clinical study of pembrolizumab monotherapy in patients with PD-L1 positive (TPS ≥1%) NSCLC

(Study 042)

The applicant has filed a partial application for the following modified indications for the treatment of

chemotherapy-naïve patients with unresectable, advanced or recurrent NSCLC, based on the pivotal data from

Studies 189, 407, and 042.

• Pembrolizumab in combination with platinum-based chemotherapy for the treatment of patients with

NSCLC (i.e., the combined target patient populations of Studies 189 [NSQ-NSCLC] and 407 [SQ-NSCLC],

regardless of PD-L1 expression status)

• Pembrolizumab monotherapy in patients with PD-L1 positive (TPS ≥1%) NSCLC

As of August 2018, the approval status of pembrolizumab outside Japan for the target patient populations of

Study 189, 407, or 042 is as follows:

• Pembrolizumab in combination with other antineoplastic drugs for the treatment of chemotherapy-naïve

patients with unresectable, advanced or recurrent NSQ-NSCLC:

Based on the pivotal data from Cohort G2) of Study 021, an approval application was filed in the US in

November 2016, and pembrolizumab received accelerated approval for the following indication in May

2017: “KEYTRUDA, in combination with pemetrexed and carboplatin, is indicated for the first-line

treatment of patients with metastatic nonsquamous NSCLC. This indication is approved under accelerated

approval based on tumor response rate and progression-free survival. Continued approval for this

indication may be contingent upon verification and description of clinical benefit in the confirmatory trials.”

Approval applications were filed based on the pivotal data from Study 189 in the US and EU in March

2018, and pembrolizumab was approved for the following indication in the US in August 2018. The

application is currently under review in the EU. Pembrolizumab, in combination with other antineoplastic

drugs, has been approved for the treatment of chemotherapy-naïve patients with unresectable, advanced or

recurrent NSQ-NSCLC in 28 countries and regions.

KEYTRUDA, in combination with pemetrexed and platinum chemotherapy, is indicated for the first-

line treatment of patients with metastatic nonsquamous NSCLC, with no EGFR or ALK genomic

tumor aberrations.

• Pembrolizumab in combination with platinum-based chemotherapy for the treatment of chemotherapy-

naïve patients with unresectable, advanced or recurrent SQ-NSCLC:

Approval applications were filed based on the pivotal data from Study 407 in the US in April 2018, and in

the EU in July 2018. The applications are currently under review. Pembrolizumab has not been approved

for the treatment in any country or region.

• Pembrolizumab monotherapy for the treatment of chemotherapy-naïve patients with unresectable,

advanced or recurrent PD-L1 positive (TPS ≥1%) NSCLC:

2) An open-label, randomized, comparative study to compare the efficacy and safety of pembrolizumab + CBDCA + PEM with those

of CBDCA + PEM alone, in chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC

6


Approval applications were filed based on the pivotal data from Study 042 in the US and EU in July 2018,

and the applications are currently under review. Pembrolizumab has not been approved for the treatment

of in any country or region.

1.2.3 MSI-High solid tumors: (a) (i) in Section 1.1. See Section 7.3 for outline of the review conducted

by PMDA.

The applicant started a global phase II study of pembrolizumab monotherapy in patient with chemotherapy-

treated patients with radically unresectable, advanced or recurrent mismatch repair deficient (dMMR) or MSI-

High (PCR) colorectal cancer (Study 164) in ** 20**, and a global phase II study of pembrolizumab

monotherapy in chemotherapy-treated patients with radically unresectable, advanced or recurrent solid tumors

(Study 158) in ** 20**. In Japan, patient enrollment in these studies was started in ** 20** and ** 20**,

respectively. The applicant has filed a partial change application for a new indication for the treatment of MSI-

High solid tumors, based on the pivotal data from Studies 164 and 158.

Pembrolizumab was selected for review under the conditional early approval system for new drugs

(PSEHB/PED Notification No. 0622-2 dated June 22, 2018).

As of August 2018, the approval status of pembrolizumab outside Japan is as follows:

• An approval application was filed based on the pivotal data from Studies 164 and 158 in the US in

September 2016, and pembrolizumab received accelerated approval for the following indication in May

2017. Pembrolizumab has been approved for the treatment of MSI-High solid tumors in 14 countries and

regions.

KEYTRUDA is indicated for the treatment of adult and pediatric patients with unresectable or

metastatic, microsatellite instability-high (MSI-H) or mismatch repair deficient solid tumors that have

progressed following prior treatment and who have no satisfactory alternative treatment options, or

colorectal cancer that has progressed following treatment with a fluoropyrimidine, oxaliplatin, and

irinotecan. This indication is approved under accelerated approval based on tumor response rate and

durability of response. Continued approval for this indication may be contingent upon verification and

description of clinical benefit in the confirmatory trials.

2. Data Relating to Quality and Outline of the Review Conducted by PMDA

Since the present application is for a new indication and a new dosage, no data relating to the quality of

pembrolizumab were submitted.

3. Non-clinical Pharmacology and Outline of the Review Conducted by PMDA

Although the present application is for a new indication and a new dosage, no new study data on non-clinical

pharmacology were submitted because the non-clinical pharmacology of pembrolizumab was evaluated during

the review process for the initial approval.

7


4. Non-clinical Pharmacokinetics and Outline of the Review Conducted by PMDA

Although the present application is for a new indication and a new dosage, no new study data on non-clinical

pharmacokinetics were submitted because the non-clinical pharmacokinetics of pembrolizumab was evaluated

during the review process for the initial approval.

5. Toxicity and Outline of the Review Conducted by PMDA

Since the present application is for a new indication and a new dosage, no data relating to the toxicity of

pembrolizumab were submitted.

6. Summary of Biopharmaceutic Studies and Associated Analytical Methods, Clinical Pharmacology,

and Outline of the Review Conducted by PMDA

6.1 Summary of biopharmaceutic studies and associated analytical methods

6.1.1 Analytical procedures

In Studies 164 and 158, (a) dMMR and (b) MSI-High (PCR) were assessed using (a) an immunohistochemistry

(IHC) assay and (b) a PCR assay, respectively. A PCR assay developed by FALCO Biosystems, the “MSI test

kit (FALCO)” was approved on September 10, 2018 as an in vitro diagnostic to assist assessment of patient

eligibility for pembrolizumab therapy.

6.2 Clinical pharmacology

6.2.1 Population pharmacokinetics (PPK) analyses

A PPK analysis was performed using a nonlinear mixed effects model, based on the pharmacokinetic data

(21,457 time points in 3881 patients) from Japanese clinical studies (Studies 011, 025, and 041), foreign clinical

studies (Studies 001, 002, 006, 037, 052, and 055), and global clinical studies (Studies 010, 024, 045, and 164)

(NONMEM version 7.2.0). The results of the PPK analysis are described below.

• The AUCss, 6wk (median [10 percentage point, 90 percentage point]) of pembrolizumab administered at 200

mg every 3 weeks (2.16 [1.45, 3.04] mg·day/mL) was similar to that at 2 mg/kg every 3 weeks (1.32 [0.722,

2.06] mg·day/mL), and lower than that at 10 mg/kg every 3 weeks (7.49 [4.32, 11.3] mg·day/mL).

• There were no clear differences in the AUCss, 6wk (median [10 percentage point, 90 percentage point]) of

pembrolizumab administered at 200 mg every 3 weeks between Japanese patients (2.15 [1.62, 3.75]

mg·day/mL) and non-Japanese patients (1.79 [1.15, 2.67] mg·day/mL).

6.R Outline of the review conducted by PMDA

Based on the data submitted, PMDA has concluded that the applicant's explanation about the clinical

pharmacology and other relevant properties of pembrolizumab are acceptable.

7. Clinical Efficacy and Safety and Outline of the Review Conducted by PMDA

7.1 Data on malignant melanoma and outline of the review conducted by PMDA

The applicant submitted efficacy and safety evaluation data, in the form of result data from a global phase III

study (Table 1).

8


Table 1. Clinical studies on the efficacy and safety

Data type Region Study

identifier Phase Subjects

No. of patients enrolled

Dosage regimen Key

endpoints

Evaluation data

Global Study 054 Ⅲ Patients with resected malignant melanoma at high risk for recurrence

1019 (a) 514 (b) 505

(a) Intravenous pembrolizumab 200 mg every 3 weeks

(b) Intravenous placebo every 3 weeks

Efficacy Safety

A summary of the clinical study is presented below. Common adverse events other than deaths reported in the

study are detailed in Section “7.4.1 Adverse events reported in clinical studies in patients with malignant

melanoma.”

7.1.1 Evaluation data

7.1.1.1 Global clinical study

7.1.1.1.1 Global phase III study (CTD 5.3.5.1.1, Study 054, ongoing since ** 20** [data cutoff date:

October 2, 2017])

A double-blind, randomized, comparative study was conducted at 134 sites in 23 countries, including Japan,

to evaluate the efficacy and safety of pembrolizumab versus placebo in patients with resected malignant

melanoma4) at high risk for recurrence3) (target sample size, 900 patients).

Patients received pembrolizumab 200 mg or placebo intravenously, every 3 weeks, for a maximum of 12

months, or until the disease recurred or a withdrawal criterion was met.

All 1019 enrolled and randomized patients (514 in the pembrolizumab group and 505 in the placebo group,

including 9 and 6 Japanese patients, respectively) were included in the intention-to-treat (ITT) population and

used for the efficacy analyses. Of these 1019 patients, 8 (5 in the pembrolizumab group and 3 in the placebo

group) did not receive the study drug. The remaining 1011 patients (509 in the pembrolizumab group and 502

in the placebo group, including 9 and 6 Japanese patients, respectively) were included in the safety analysis

set.

The primary endpoint was recurrence-free survival (RFS)5) as assessed by the investigator. At the start of the

study, the primary analysis was planned to be conducted in the overall study population and the PD-L1 positive

(melanoma [MEL] score ≥2) subpopulation when a total of 409 RFS events had occurred; however,

******************************************************* with respect to ********************

in ****************,6) and it was considered that *********************************************

based on *************************, the study protocol was amended to perform an interim analysis on

the overall study population to assess the superiority of pembrolizumab to placebo when a total of 330 RFS

4) Patients whose tumors were completely resected 3) Patients with (a) stage IIIA (>1 mm lymph node metastasis), (b) stage IIIB, or (c) stage IIIC melanoma were defined as being at a

high risk for recurrence. 5) RFS was defined as the time between the date of randomization and the date of first recurrence (local, regional, or distant metastasis)

or death, whichever occurs first. 6) *****************************************************************************************************：：

*********************************************************************************

9


events have occurred, and the final analysis when a total of 409 RFS events have occurred. In addition, if the

primary RFS analysis showed a statistically significant result, each secondary endpoint was to be analyzed

sequentially, starting with distant metastasis-free survival (DMFS), followed by overall survival (OS) (Protocol

Amendment Version 2, dated October 2, 2017). To identify PD-L1 positive (MEL score ≥2) patients, PD-L1

levels in tumor tissue samples were determined using the “PD-L1 IHC 22C3 pharmDx ‘Dako’,” by Dako Japan

Co., Ltd.

In the analysis of the overall study population, a one-sided significance level of 0.014 was used. In the analysis

of the PD-L1 positive (MEL score ≥2) subpopulation, the one-sided significance level was determined by the

method proposed by Spiessens and Debois (Contemp Clin Trials 2010;31:647-56). If a statistically significance

was found in either the overall study population or the PD-L1 positive subpopulation, another population was

analyzed at a one-sided significance level of 0.025. A Lan-DeMets α spending function of the O’Brien-Fleming

type was used to control the probability of a type I error associated with the interim analysis.

The interim analysis (data cutoff date: October 2, 2017) was performed when a total of 351 RFS events had

occurred. The results of the interim RFS analysis in the primary analysis sets (the overall study population and

the PD-L1 positive [MEL score ≥2] subpopulation) (Table 2) and their Kaplan-Meier curves of RFS (Figures

1 and 2) are shown below.

Table 2. Interim RFS analysis (Investigator assessment [data cutoff date: October 2, 2017])

Overall PD-L1 positive (MEL score ≥2)

Pembrolizumab Placebo Pembrolizumab Placebo

N 514 505 428 425 Number of events (%) 135 (26.3) 216 (42.8) 102 (23.8) 176 (41.4)

Median [95% CI] (months) — [—, —] 20.4 [16.2, —] —[—, —] — [17.1, —] Hazard ratio [CI]*1 0.57 [0.43, 0.74]*2 0.54 [0.42, 0.69]*3

P value (one-sided)*4 <0.0001*5 <0.0001*6

—, Not reached; *1, Cox regression stratified by disease stage (IIIA, IIIB, IIIC [1 to 3 lymph node metastases present], or IIIC [4 or more lymph node metastases present]); *2, 98.4% CI; *3, 95% CI; *4, Log-rank test stratified by disease stage (IIIA, IIIB, IIIC [1 to 3 lymph node metastases present], or IIIC [4 or more lymph node metastases present]); *5, One-sided significance level = 0.008; *6, One-sided significance level = 0.0155

10


Pembrolizumab

Placebo

Placebo

Figure 1. Kaplan Meier curves for the interim RFS analysis (Investigator assessment, overall study population [data cutoff date: October 2, 2017])

Pembrolizumab

Placebo

Figure 2. Kaplan Meier curves for the interim RFS analysis (Investigator assessment, PD-L1 positive subpopulation [MEL score ≥2] [data cutoff date: October 2, 2017])

Pembrolizumab

Placebo

Pembrolizumab

11


The safety analysis indicated that 1 of 509 patients (0.2%) died in the pembrolizumab group during the

treatment period or within 90 days after the last dose. No Japanese patients died. The cause of death was a drug

reaction with eosinophilia and systemic symptoms, for which a causal relationship to the study drug was denied.

7.1.R Outline of the review conducted by PMDA on malignant melanoma

7.1.R.1 Efficacy

PMDA evaluated the efficacy of pembrolizumab in Japanese patients, in terms of the consistency between the

overall study population and the Japanese subpopulation of Study 054, according to “Basic Principles on

Global Clinical Trials” (PFSB/ELD Notification No. 0928010 dated September 28, 2007) and “Basic Principles

on Global Clinical Trials (Reference Cases)” (Administrative Notice dated September 5, 2012).

Based on the discussion presented below, PMDA concluded that the efficacy of pembrolizumab was

demonstrated in patients with resected malignant melanoma at high risk for recurrence.

7.1.R.1.1 Selection of control group

When Study 054 was planned, the National Comprehensive Cancer Network Clinical Practice Guidelines

(NCCN guidelines) (malignant melanoma) (v.3.2015) recommended no standard treatment regimen for the

target patient population of the study; therefore, placebo was selected as a control.

PMDA accepted the applicant's explanation.

7.1.R.1.2 Efficacy endpoints

The applicant's rationale for the selection of RFS as the primary endpoint in Study 054:

In Study 054, an RFS event was defined as either: (a) a local recurrence, a regional lymph node metastasis, or

a distant metastasis; or, (b) death. The prolongation of RFS, which includes (a) and (b) as defined above, is

clinically meaningful in patients with malignant melanoma who have undergone resection, because delayed

recurrence would lead to the long-term maintenance of patient physical functions and quality of life. Therefore,

the selection of RFS as the primary endpoint in Study 054 is appropriate.

PMDA's view:

The primary endpoint in Study 054 should have been OS, because treatment for patients with resected

malignant melanoma is usually intended to prolong their survival. However, the applicant's explanation that

prolonged RFS is clinically meaningful in such patients is understandable. In view of the applicant's

explanation, and for other reasons, PMDA has concluded that evaluating the efficacy of pembrolizumab based

on RFS data is acceptable.

7.1.R.1.3 Efficacy evaluation results

The primary analysis in Study 054 demonstrated the superiority of pembrolizumab to placebo in investigator-

assessed RFS (the primary endpoint) [see Section 7.2.1.1.1]. The superiority of pembrolizumab to placebo was

12


also demonstrated in the PD-L1 positive (MEL score ≥2) subpopulation [see Section 7.1.1.1.1]. Table 3

compares the results of investigator-assessed RFS between the PD-L1 positive (MEL score ≥2) and negative

(MEL score ≤1) subpopulations, and Figure 3 shows the Kaplan-Meier curves of RFS in the PD-L1 negative

(MEL score ≤1) subpopulation.

Table 3. Efficacy by MEL score in tumor tissue samples (Study 054)

PD-L1 expression

Treatment N RFS

Median [95% CI] (months)

Hazard ratio* [95% CI]

P value for interaction

MEL score ≤1 Pembrolizumab 59 — [15.8, —]

0.47 [0.26, 0.85] 0.1028

Placebo 57 19.4 [5.7, —]

MEL score ≥2 Pembrolizumab 428 — [—, —]

0.54 [0.42, 0.69] Placebo 425 — [17.1, —]

—, Not reached; *, Cox regression stratified by disease stage (IIIA, IIIB, IIIC [1 to 3 lymph node metastases present], or IIIC [4 or more lymph node metastases present])

Figure 3. Kaplan-Meier curves for the interim RFS analysis in the PD-L1 negative (MEL score ≤1)

subpopulation (data cutoff date: October 2, 2017)

Table 4 shows the results of the interim RFS analysis in Japanese patients in the primary analysis sets (the

overall study population and the PD-L1 positive [MEL score ≥2] subpopulation) of Study 054, and Figures 4

and 5 show the corresponding Kaplan-Meier curves for RFS.

Pembrolizumab

Placebo

Pembrolizumab

Placebo

13


Table 4. Interim RFS analysis in Japanese patients (Investigator assessment, primary analysis sets [data cutoff date: October 2, 2017])

Overall PD-L1 positive (MEL score ≥2)

Pembrolizumab Placebo Pembrolizumab Placebo

N 9 6 9 3 Number of events (%) 3 (33.3) 3 (50.0) 3 (33.3) 2 (66.7)

Median [95% CI] (months) — [1.4, —] — [2.8,—] — [1.4,—] 8.3 [2.8,—] Hazard ratio [CI]*1 0.64 [0.09, 4.59]*2 0.44 [0.07, 2.63]*3

P value (one-sided)*4 0.29218 0.18294 —, Not reached; *1, Non-stratified Cox regression; *2, 98.4% CI; *3, 95% CI; *4, Non-stratified log-rank test

Figure 4. Kaplan-Meier curves for the interim RFS analysis in Japanese patients

(Investigator assessment, overall study population [data cutoff date: October 2, 2017])

Pembrolizumab

Placebo

Pembrolizumab

Placebo

14


Figure 5. Kaplan-Meier curves for the interim RFS analysis in Japanese patients

(Investigator assessment, PD-L1 positive (MEL score ≥2) subpopulation [data cutoff date: October 2, 2017])

PMDA's view:

Considering that pembrolizumab was demonstrated to prolong OS in an open-label, randomized, foreign phase

III study in patients with unresectable malignant melanoma (Study 006) (see “Review Report for Keytruda

Injection 20 mg, Keytruda Injection 100 mg dated August 30, 2016”), PMDA has concluded that the efficacy

of pembrolizumab has been demonstrated in patients with resected malignant melanoma at high risk for

recurrence, in view of the following findings.

• In Study 054, pembrolizumab was superior to placebo in investigator-assessed RFS (the primary endpoint),

and the observed RFS prolongation was clinically meaningful.

• Although the sample size of the Japanese subpopulation in Study 054 and the number of events occurring

in the subpopulation were limited and not sufficient to appropriately evaluate the efficacy of

pembrolizumab in Japanese patients, based on the results from the Japanese subpopulation, the RFS in the

Japanese subpopulation did not tend to clearly differ from that in the overall study population.

7.1.R.2 Safety [For adverse events, see “7.4.1 Adverse events, etc. reported in clinical studies in patients

with malignant melanoma.”]

PMDA's view:

As a result of its review described in the section below, PMDA concluded that the adverse events that require

special attention after the administration of pembrolizumab to patients with resected malignant melanoma are

Pembrolizumab

Placebo

Pembrolizumab

Placebo

15


those identified as requiring attention at the regulatory reviews for the approved indications7) and should also

be closely monitored when administrating pembrolizumab.

PMDA's conclusion:

Although attention should be paid to the above events, pembrolizumab was concluded to be tolerable in patients

with resected malignant melanoma, as long as they are followed up by physicians with sufficient knowledge

and experience in cancer chemotherapy, through monitoring of adverse events, differential diagnosis and

patient management in anticipation of adverse reactions due to an excessive immune response, drug

interruption, or other appropriate actions.

7.1.R.2.1 Differences in the safety profile between Japanese and non-Japanese patients

The applicant's explanation about the safety profile of pembrolizumab in patients with resected malignant

melanoma, based on the safety data from Study 054:

Table 5 shows a summary of the safety data from Study 054.

Table 5. Safety summary (Study 054)

n (%)

Pembrolizumab N = 509

Placebo N = 502

All adverse events 475 (93.3) 453 (90.2) Grade ≥3 adverse events 158 (31.0) 96 (19.1) Serious adverse events 128 (25.1) 82 (16.3) Adverse events leading to death 1 (0.2) 0 Adverse events leading to drug discontinuation 70 (13.8) 18 (3.6) Adverse events leading to drug interruption 96 (18.9) 46 (9.2)

In Study 054, adverse events of any grade reported with a ≥5% higher incidence in the pembrolizumab group

than in the placebo group were pruritus (19.4% [99 of 509 patients] in the pembrolizumab group, 11.6% [58

of 502 patients] in the placebo group), hypothyroidism (14.7% [75 of 509 patients], 2.8% [14 of 502 patients]),

and hyperthyroidism (10.4% [53 of 509 patients], 1.2% [6 of 502 patients]). No Grade ≥3 adverse events,

serious adverse events, adverse events leading to death, adverse events leading to drug discontinuation, or

adverse events leading to drug interruption were reported with a ≥2% higher incidence in the pembrolizumab

group than in the placebo group.

The incidences of adverse events of any grade by PD-L1 expression status in the pembrolizumab group were

89.8% in PD-L1 negative (MEL score ≤1) patients and 94.1% in PD-L1 positive (MEL score ≥2) patients,

those of Grade ≥3 adverse events were 32.2% and 30.7%, and those of serious adverse events were 18.6% and

7) Gastrointestinal disorders, skin disorders, neurological disorders, hepatic function disorder, cholangitis sclerosing, eye disorders,

endocrine disorders, renal impairment, interstitial lung disease (ILD), infusion-related reaction (IRR), pancreatitis, myositis, rhabdomyolysis, encephalitis and meningitis, myasthenia gravis, myocarditis, immune thrombocytopenic purpura, haemolytic anaemia, and aplasia pure red cell (see “Review Report for “Keytruda Injection 20 mg, Keytruda Injection 100 mg dated November 15, 2016”).

16


27.0%. These results suggested that pembrolizumab was tolerable, regardless of PD-L1 expression status, in

patients with resected malignant melanoma.

The applicant's explanation about the differences in the safety profile of pembrolizumab between patients with

resected malignant melanoma and those treated for the approved applications:

Table 6 shows a comparison of the incidences of adverse events in clinical studies (a) to (i) below.

(a) A global phase III study in resected malignant melanoma (Study 054)

(b) A foreign phase II study (Study 002) and a foreign phase III study (Study 006) in patients with unresectable

malignant melanoma

(c) A global phase II/III study in platinum-based chemotherapy-treated patients with unresectable, advanced

or recurrent PD-L1 positive (TPS ≥1%) NSCLC (Study 010), and a global phase III study in

chemotherapy-naïve patients with unresectable, advanced or recurrent PD-L1 positive (TPS ≥50%)

NSCLC (Study 024)

(d) A global phase III study in chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-

NSCLC (Study 189)

(e) A global phase III study in chemotherapy-naïve patients with unresectable, advanced or recurrent SQ-

NSCLC (Study 407)

(f) A global phase III study in chemotherapy-naïve patients with unresectable, advanced or recurrent PD-L1

positive (TPS ≥1%) NSCLC (Study 042)

(g) A global phase II study in MSI-High colorectal cancer (Study 164 [Cohort A])

(h) A global phase II study in patients with cHL (Study 087)

(i) A global phase III study in patients with radically unresectable urothelial carcinoma that has progressed

after cancer chemotherapy (Study 045)

Table 6. Safety summary by cancer type*1

n (%) (a) (b) (c) (d) (e) (f) (g) (h) (i) N = 509 N = 912 N = 836 N = 405 N = 278 N = 636 N = 61 N = 210 N = 266

All adverse events 475 891 809 404 273 610 60 202 248 (93.3) (97.7) (96.8) (99.8) (98.2) (95.9) (98.4) (96.2) (93.2) Grade ≥3 adverse events

158 425 396 272 194 318 36 53 139 (31.0) (46.6) (47.4) (67.2) (69.8) (50.0) (59.0) (25.2) (52.3)

Adverse events leading to death

1 39 52 27 23 70 1 2 13 (0.2) (4.3) (6.2) (6.7) (8.3) (11.0) (1.6) (1.0) (4.9)

Serious adverse events

128 341 314 202 113 259 29 34 104 (25.1) (37.4) (37.6) (49.9) (40.6) (40.7) (47.5) (16.2) (39.1)

Adverse events leading to drug discontinuation*2

70 (13.8)

115 (12.6)

68 (8.1)

112 (27.7)

65 (23.4)

122 (19.2)

4 (6.6)

11 (5.2)

22 (8.3)

Adverse events leading to drug interruption*3

96 213 209 223 164 212 17 54 54

(18.9) (23.4) (25.0) (55.1) (59.0) (33.3) (27.9) (25.7) (20.3)

*1, Pembrolizumab was administered at: an intravenous dose of 2 mg/kg every 3 weeks, 10 mg/kg every 2 weeks, or 10 mg/kg every 3 weeks in patients with unresectable malignant melanoma; an intravenous dose of 2 mg/kg every 3 weeks, 10 mg/kg every 3 weeks, or 200 mg every 3 weeks in patients with PD-L1 positive NSCLC; or an intravenous dose of 200 mg every 3 weeks in other patients; *2, Adverse events that led to the discontinuation of pembrolizumab or the concomitant chemotherapy in chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC or SQ-NSCLC; or adverse events that led to the discontinuation of pembrolizumab in other patients; *3, Adverse events that led to the interruption of pembrolizumab or the concomitant chemotherapy

17


in chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC or SQ-NSCLC; or adverse events that led to the interruption of pembrolizumab in other patients

Adverse events of any grade reported with a ≥5% higher incidence in patients with resected malignant

melanoma than in those with any other type of cancer were hypertension (14.5% in patients with resected

malignant melanoma, 3.9% in patients with unresectable malignant melanoma, 3.7% in patients with PD-L1

positive NSCLC, 1.9% in patients with cHL, 4.1% in patients with radically unresectable urothelial carcinoma

that has progressed after cancer chemotherapy), weight increased (12.4%, 2.2%, 1.1%, 3.8%, 0.8%), aplasia

influenza like illness (10.8%, 5.5%, 2.5%, 1.9%, 3.4%), and hyperthyroidism (10.4%, 4.2%, 5.3%, 2.9%, 3.8%).

The Grade ≥3 adverse event reported with a ≥2% higher incidence in patients with resected malignant

melanoma than in those with any other type of cancer was hypertension (5.5%, 1.4%, 1.4%, 0.5%, 2.3%). The

serious adverse event reported with a ≥2% higher incidence in patients with resected malignant melanoma than

in those with any other type of cancer was basal cell carcinoma (3.3%, 0.9%, 0.1%, 0%, 0%). No adverse

events led to death, drug discontinuation, or interruption, with a ≥2% higher incidence in patients with resected

malignant melanoma than in those with any other type of cancer.

Adverse events of any grade, which did not occur in patients treated with pembrolizumab for the approved

indications but reported in ≥3 patients in the pembrolizumab group of Study 054 were anal haemorrhage, oral

lichen planus, tenderness, and sensation of pharynx strangled in 3 patients each. There were no Grade ≥3

adverse events, serious adverse events, or adverse events leading to death, drug discontinuation, or interruption,

which did not occur in patients treated for the approved indications but were reported in ≥2 patients in the

pembrolizumab group of Study 054.

As was described above, some adverse events were more frequently reported in patients with resected

malignant melanoma than in patients treated for the approved indications, or identified newly in patients with

resected malignant melanoma. However, no clear differences were noticed in the incidences of serious adverse

events and other notable adverse events between these patient populations. Therefore, the safety of

pembrolizumab is comparable in patients with resected malignant melanoma and those treated for the approved

indications.

The applicant's explanation about the differences in the safety of pembrolizumab between Japanese patients

and non-Japanese patients, based on the safety data from Study 054:

In Study 054, adverse events of any grade reported in ≥2 Japanese patients were diarrhoea (4 Japanese patients

[44.4%], 137 non-Japanese patients [27.4%]), rash (3 patients [33.3%], 64 patients [12.8%]), nasopharyngitis

(2 patients [22.2%], 42 patients [8.4%]), upper respiratory tract infection (2 patients [22.2%], 37 patients

[7.4%]), ALT increased (2 patients [22.2%], 35 patients [7.0%]), decreased appetite (2 patients [22.2%], 34

patients [6.8%]), AST increased (2 patients [22.2%], 25 patients [5.0%]), abdominal pain upper (2 patients

[22.2%], 21 patients [4.2%]), and eczema (2 patients [22.2%], 18 patients [3.6%]). Serious adverse events

reported in Japanese patients were diarrhoea (1 patient [11.1%], 4 patients [0.8%]), pyrexia (1 patient [11.1%],

3 patients [0.6%]), type 1 diabetes mellitus (1 patient [11.1%], 2 patients [0.4%]), decreased appetite (1 patient

[11.1%], 1 patient [0.2%]), and rash (1 patient [11.1%], 0 patients). Grade ≥3 adverse events reported in

18


Japanese patients were type 1 diabetes mellitus (1 patient [11.1%], 4 patients [0.8%]) and rash (1 patient

[11.1%], 0 patients). The adverse events that led to drug discontinuation in Japanese patients was diarrhoea (1

patient [11.1%], 4 patients [0.8%], and adverse events that led to drug interruption in Japanese patients were

diarrhoea (1 patient [11.1%], 11 patient [2.2%]), AST increased (1 patient [11.1%], 6 patients [1.2%]), pyrexia

(1 patient [11.1%], 1 patient [0.2%]), type 1 diabetes mellitus (1 patient [11.1%], 1 patient [0.2%]), cough (1

patient [11.1%], 3 patients [0.6%]), and rash (1 patient [11.1%], 0 patients). No Japanese patients died due to

adverse events.

As only limited clinical experience with pembrolizumab is available in Japanese patients with resected

malignant melanoma, it is difficult to assess the differences in the safety of pembrolizumab between Japanese

and non-Japanese patients based on the results of Study 054. Nevertheless, no adverse events would require

special attention in Japanese patients, because no serious adverse events were reported in ≥2 Japanese patients

in Study 054, and due to other findings.

PMDA's view:

Although several adverse events were reported with a higher incidence in patients with resected malignant

melanoma than in those treated for the approved indications, or were newly identified in patients with resected

malignant melanoma, most of these adverse events were of Grade ≤2. PMDA has therefore concluded that

pembrolizumab would be tolerable in patients with resected malignant melanoma, as long as they are followed

up by physicians with sufficient knowledge and experience in cancer chemotherapy, through monitoring of

adverse events, differential diagnosis, and patient management in anticipation of adverse reactions caused by

an excessive immune response, drug interruption, or other appropriate actions.

PMDA has also concluded that the currently available data indicate no adverse events requiring special

attention in Japanese patients with resected malignant melanoma, although clinical experience with

pembrolizumab in Japanese patients is limited.

7.1.R.3 Clinical positioning and indications

In the present partial change application, the applicant has proposed to change the approved indication of

“unresectable malignant melanoma” to “malignant melanoma,” and to delete the following precautionary

statement included in the approved “Precautions for Indications” section of the package insert: “The efficacy

and safety of pembrolizumab in adjuvant chemotherapy have not been established.”

As a result of its review described in Sections “7.1.R.1 Efficacy,” “7.1.R.2 Safety,” and the section below,

PMDA concluded that the indication of pembrolizumab should be “malignant melanoma,” as proposed by the

applicant, provided that the disease stage and other characteristics of the patients enrolled in Study 054 are

detailed in the “Clinical Studies” section of the package insert, and that the following statement is included in

the “Precautions for Indications” section.

• Appropriate patients should be selected based on a good understanding of the efficacy and safety of

pembrolizumab, after thoroughly understanding the “Clinical Studies” section.

19


7.1.R.3.1 Clinical positioning and indication of pembrolizumab

In clinical practice guidelines or representative textbooks for clinical oncology in and outside Japan,

pembrolizumab therapy for resected malignant melanoma is currently described as follows:

[Clinical practice guidelines]

• US National Cancer Institute Physician Data Query (NCI-PDQ) (version dated July 19, 2018):

The results of Study 054 demonstrated the efficacy of pembrolizumab in the treatment of resected

malignant melanoma.

PMDA asked the applicant to explain about the clinical positioning and indication of pembrolizumab in the

treatment of resected malignant melanoma.

The applicant's explanation:

Since the results of Study 054 indicated that pembrolizumab can be positioned as an adjuvant chemotherapy

option for patients with resected malignant melanoma at high risk for recurrence, the approved indication of

“unresectable malignant melanoma” should be changed to “malignant melanoma.” In addition, the following

precautionary statement that was included in the “Precautions for Indications” section of the package insert at

the previous approval should be deleted: “The efficacy and safety of pembrolizumab in adjuvant chemotherapy

have not been established.”

Administration of pembrolizumab to patients with different stages of malignant melanoma who were ineligible

for entry in Study 054 should not be recommended because no clinical data is available as to the clinical benefit

of pembrolizumab in such patients. However, in view of the fact that pembrolizumab is intended to be used by

physicians with sufficient knowledge and experience in cancer chemotherapy, there is little necessity to specify

the intended disease stage in the “Indications” section; thus, the disease stages of the patients enrolled in Study

054 will be stated in the “Clinical Studies” section.

The applicant's explanation about the choice between pembrolizumab and conventional antineoplastic drugs in

the treatment of resected malignant melanoma:

Pembrolizumab and conventional antineoplastic drugs should be chosen according to BRAF mutation status as

described below.

• BRAF mutation-negative malignant melanoma (conventional antineoplastic drug: nivolumab):

There have been no clinical data comparing the efficacy and safety of pembrolizumab with those of

nivolumab; therefore, it is unclear which drug should be chosen first.

• BRAF mutation-positive malignant melanoma (conventional antineoplastic drugs: dabrafenib and

trametinib):

There have been no clinical data comparing the efficacy and safety of pembrolizumab with those of

dabrafenib or trametinib; therefore, it is unclear which drug should be chosen first.

20


PMDA's view:

PMDA accepted the applicant's explanation in general and has concluded that the indication of pembrolizumab

should be “malignant melanoma,” as proposed by the applicant. However, since there have been no clinical

data demonstrating the clinical benefit of pembrolizumab in patients with different stages of malignant

melanoma who were ineligible for entry in Study 054, and for other reasons, the following precautionary

statement should be included in the “Precautions for Indications” section of the package insert, provided that

the disease stages of the patients enrolled in Study 054 are detailed in the “Clinical Studies” section.


pembrolizumab, after carefully reading the “Clinical Studies” section.

7.1.R.4 Dosage and administration

The approved dosage and administration for the approved indication of “unresectable malignant melanoma” is

a body weight-based dose (2 mg/kg every 3 weeks), while a fixed-based dose (200 mg every 3 weeks) has been

approved for other approved indications.

The present partial change application has proposed the following dosage and administration for both

indications of malignant melanoma (treatment of unresectable malignant melanoma and the adjuvant treatment

of malignant melanoma): “The usual adult dosage is 200 mg of pembrolizumab (genetical recombination)

infused intravenously over 30 minutes every 3 weeks,” as well as the following statements to be included in

the “Precautions for Dosage and Administration” section of the package insert (unchanged from the previous

approval).

• The efficacy and safety of pembrolizumab in combination with other antineoplastic drugs have not been

established.

• Criteria for interruption and discontinuation of pembrolizumab in case of adverse reactions


PMDA concluded that the dosage and administration of pembrolizumab for malignant melanoma should be

“The usual adult dosage is 200 mg of pembrolizumab (genetical recombination) infused intravenously over 30

minutes every 3 weeks. For the adjuvant treatment of malignant melanoma, the maximum duration of treatment

should be 12 months.” provided that the statements proposed by the applicant (unchanged from the approved

statements) are included in the “Precautions for Dosage and Administration” section of the package insert.

7.1.R.4.1 Dosage for the treatment of unresectable malignant melanoma


The dosage and administration of pembrolizumab for the treatment of unresectable malignant melanoma can

be changed from the approved body weight-based dose (2 mg/kg every 3 weeks) to a fixed-based dose (200

mg every 3 weeks), based on the following findings, as well as the results of a PPK analysis [see Section 6.2.1].

• The pharmacokinetics of pembrolizumab was suggested to show similar inter-individual variability with

the body weight-based dose and the fixed-based dose (J Immunother Cancer 2017;5:43).

21


• The results of foreign clinical studies (Studies 001, 002, and 006) showed no clear association between the

AUCss,6wk of pembrolizumab and its efficacy or safety, over the investigated dosage range (2 mg/kg every

3 weeks, 10 mg/kg every 2 weeks, and 10 mg/kg every 3 weeks) (see “Review Report of Keytruda Injection

20 mg, Keytruda Injection 100 mg dated August 30, 2016”). Based on the above, and other findings, the

change of dosage and administration from 2 mg/kg every 3 weeks to 200 mg every 3 weeks is unlikely to

result in the reduced efficacy of pembrolizumab.

• No clear differences were found in the safety of pembrolizumab administered at 2 mg/kg every 3 weeks

among cancer types [see Sections 7.1.R.2, 7.2.R.2, and 7.3.R.1]. Likewise, pembrolizumab administered

at a dose of 200 mg every 3 weeks is unlikely to cause clear differences in the incidence of adverse events

among cancer types.


7.1.R.4.2 Dosage and administration of pembrolizumab for the adjuvant treatment of malignant

melanoma

The applicant's explanation about the dosage and administration of pembrolizumab for the adjuvant treatment

of malignant melanoma:

In Study 054, based on the results of PK/PD analyses (see “Review Report of Keytruda Injection 20 mg,

Keytruda Injection 100 mg, dated August 30, 2016” and “Review Report of Keytruda Injection 20 mg,

Keytruda Injection 100 mg, dated November 15, 2016”) and other findings, a fixed-based dose of 200 mg every

3 weeks, rather than a body weight-based dose, was selected as the dosage and administration of

pembrolizumab. As a result, the study demonstrated the clinically significant efficacy [see Section 7.1.R.1] and

acceptable tolerability [see Section 7.1.R.2] of pembrolizumab. Therefore, the dosage and administration of

pembrolizumab for the adjuvant treatment of malignant melanoma has been proposed as, “The usual adult

dosage is 200 mg of pembrolizumab (genetical recombination) infused intravenously over 30 minutes every 3

weeks.”

In Study 054, the maximum duration of treatment with pembrolizumab was 12 months. PMDA asked the

applicant to explain the necessity of specifying the duration of pembrolizumab therapy for resected malignant

melanoma.


No clinical data are available in patients who have been treated with pembrolizumab for resected malignant

melanoma beyond 12 months. However, the treatment duration of pembrolizumab of up to 12 months in Study

054 will be stated in the “Clinical Studies” section of the package insert; therefore, the duration of

pembrolizumab therapy does not have to be limited to 12 months in the “Dosage and Administration” section.

22


PMDA's view:

PMDA accepted the applicant's explanation in general. However, the duration of adjuvant treatment with

pembrolizumab for malignant melanoma should be specified in the “Dosage and Administration” section of

the package insert, in view of the following facts.

• In Study 054, which involved patients with resected malignant melanoma, the maximum duration of

treatment with pembrolizumab was 12 months; therefore, no clinical data are available as to the clinical

benefit of pembrolizumab administered beyond 12 months.

• The patients enrolled in Study 054 underwent surgery intended for complete cure. Administration of

pembrolizumab without careful consideration should be avoided in patients who have completed a 12-

month adjuvant treatment with pembrolizumab.

7.1.R.5 Post-marketing investigations

Post-marketing investigations in patients with malignant melanoma will be described in Section 7.3.R.4,

together with those in patients with NSCLC or MSI-High solid tumors.

7.2 Data on NSCLC and outline of the review conducted by PMDA

The applicant submitted efficacy and safety evaluation data, in the form of result data from 1 Japanese phase I

study, 3 global phase III studies, and 1 foreign phase I/II study (5 studies in total) (Table 7).

23


Table 7. Clinical studies on efficacy and safety Data type

Region Study



Brief description of dosage regimen* Endpoints

Evaluation data

Japan Study 011

I

Chemotherapy-naïve patients with unresectable, advanced or recurrent NSCLC Part B: Patients with NSQ-NSCLC Part C: Patients with SQ-NSCLC

Part B Cohort 1:

6 Cohort 2:

6

Part C Cohort 1:

8 Cohort 2:

6

Part B Cohort 1: Intravenous pembrolizumab 200 mg in combination with CDDP + PEM for 4 cycles, followed by intravenous pembrolizumab and PEM Cohort 2: Intravenous pembrolizumab 200 mg in combination with CBDCA + PEM for 4 cycles, followed by intravenous pembrolizumab and PEM Part C Cohort 1: Intravenous pembrolizumab 200 mg in combination with CBDCA + PTX for 4 cycles, followed by intravenous pembrolizumab Cohort 2: Intravenous pembrolizumab 200 mg in combination with CBDCA + nab-PTX for 4 cycles, followed by intravenous pembrolizumab

Safety Tolerability

PK

Global

Study 189

Ⅲ

Chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC

616 (a) 410 (b) 206

In combination with CDDP/PEM or CBDCA/PEM,(a) Intravenous pembrolizumab 200 mg for 4 cycles,

followed by intravenous pembrolizumab and PEM

(b) Intravenous placebo for 4 cycles, followed by intravenous placebo and PEM

Efficacy Safety

Study 407

Ⅲ

Chemotherapy-naïve patients with unresectable, advanced or recurrent SQ-NSCLC

559 (a) 278 (b) 281

In combination with CBDCA + PTX or CBDCA + nab-PTX, (a) Intravenous pembrolizumab 200 mg for 4 cycles,

followed by intravenous pembrolizumab (b) Intravenous placebo for 4 doses, followed by

intravenous placebo

Efficacy Safety

Study 042

Ⅲ

Chemotherapy-naïve patients with unresectable, advanced or recurrent PD-L1 positive (TPS ≥1%) NSCLC

1275 (a) 638 (b) 637

(a) Intravenous pembrolizumab 200 mg (b) Intravenous CBDCA + PEM or CBDCA + PTX

for 6 cycles, followed by intravenous PEM

Efficacy Safety

Foreign Study 021

I/II

Phase I part Cohort A: Chemotherapy-naïve patients with advanced or recurrent NSCLC Cohort C: Chemotherapy-naïve patients with advanced or recurrent NSQ-NSCLC Phase II part Cohort G: Chemotherapy-naïve patients with advanced or recurrent NSQ-NSCLC

Phase I part Cohort A:

25

Cohort C: 24

Phase II part Cohort G:

(a) 60 (b) 63

Phase I part Cohort A: Intravenous pembrolizumab 2 or 10 mg/kg in combination with CBDCA + PTX for 4 cycles, followed by intravenous pembrolizumab Cohort C: Intravenous pembrolizumab 2 or 10 mg/kg in combination with CBDCA + PEM for 4 cycles, followed by intravenous pembrolizumab and PEM Phase II part Cohort G: (a) Intravenous pembrolizumab 200 mg in

combination with CBDCA + PEM for 4 cycles, followed by intravenous pembrolizumab and PEM

(b) Intravenous CBDCA + PEM for 4 cycles, followed by intravenous PEM

Efficacy Safety

* CBDCA was administered intravenously on Day 1 of each 3-week cycle: at a dose producing AUC of 5 mg·min/mL in Part B of Study 011, Cohorts C and G of Study 021, and Study 189; at a dose producing AUC of 6 mg·min/mL in Part C of Study 011, Cohort A of Study 021, and Study 407; or, at a dose producing AUC of 5 or 6 mg·min/mL in Study 042. CDDP was administered intravenously at a dose of 75 mg/m2, PEM at a dose of 500 mg/m2, PTX at a dose of 175 or 200 mg/m2, and nab-PTX at 100 mg/m2. A summary of the clinical studies is shown below. Major adverse events, excluding deaths, were reported in

the studies are detailed in Section “7.4.2 Adverse events reported in clinical studies in patients with NSCLC.”

24



7.2.1.1 Japanese clinical study

7.2.1.1.1 Japanese phase I study (CTD 5.3.5.2.1 and 5.3.5.2.2, Study 011, Parts B and C, ongoing since

** 20** [data cutoff date: Part B, ** **, 20**; Part C, ** **, 20**])

An open-label, uncontrolled study was conducted at 4 sites in Japan to evaluate the safety, tolerability, and

other aspects of pembrolizumab in combination with platinum-based chemotherapy, in chemotherapy-naïve

patients with unresectable, advanced or recurrent NSQ-NSCLC (Part B; target sample size, 12 to 18 patients)

or SQ-NSCLC (Part C; target sample size, 12 to 18 patients).

As described below, the treatment was continued until disease progression was observed or a withdrawal

criterion was met.

• Part B:

Cohort 1: Patients received an intravenous dose of pembrolizumab 200 mg in combination with CDDP

75 mg/m2 and PEM 500 mg/m2 on Day 1 of each 3-week cycle for 4 cycles, followed by an intravenous

dose of pembrolizumab in combination with PEM for a total of 24 months.

Cohort 2: Patients received an intravenous dose of pembrolizumab 200 mg in combination with

CBDCA (AUC 5 mg·mL/min) and PEM 500 mg/m2 on Day 1 of each 3-week cycle for 4 cycles,

followed by an intravenous dose of pembrolizumab in combination with PEM for a total of 24 months.

• Part C:


CBDCA (AUC 6 mg·mL/min) and PTX 200 mg/m2 on Day 1 of each 3-week cycle for 4 cycles,

followed by an intravenous dose of pembrolizumab for a total of 24 months.


CBDCA (AUC 6 mg·mL/min) and nab-PTX 100 mg/m2 on Day 1 of each 3-week cycle for 4 cycles,

followed by an intravenous dose of pembrolizumab for a total of 24 months.

All 26 enrolled patients (Part B [6 in Cohort 1 and 6 in Cohort 2], Part C [8 in Cohort 1 and 6 in Cohort 2])

were treated with the study drug and included in the safety analysis set.

During the first 21 days after the start of treatment with the study drug, dose limiting toxicities (DLTs) were

evaluated. Although DLTs were found in 1 of 6 patients in Cohort 1 of Part B (Grade 4 hyponatraemia) and 2

of 6 patients in Cohort 1 of Part C (Grade 3 febrile neutropenia in both patients), pembrolizumab administered

in combination with platinum-based chemotherapy was tolerable.

The safety analysis indicated that 2 of 12 patients (16.7%) in Part B (2 of 6 patients [33.3%] in Cohort 2) and

1 of 14 patients (7.1%) in Part C (1 of 6 patients [16.7%] in Cohort 2) died during the treatment period or

within 90 days after the last dose. The causes of death were ILD and pneumonitis in 1 patient each in Part B,

and lung infection in 1 patient in Part C. A causal relationship to the study drug was not ruled out for the ILD

and pneumonitis in 1 patient each in Part B.

25



7.2.1.2.1 Global phase III study (CTD 5.3.5.1.1, Study 189, ongoing since February 2016 [data cutoff

date: November 8, 2017])

A double-blind, randomized, comparative study was conducted at 143 sites in 16 countries and regions,

including Japan, to evaluate the efficacy and safety of pembrolizumab versus placebo, in combination with a

platinum + PEM in chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC (target

sample size, 570 subjects).

Patients received an intravenous dose of pembrolizumab 200 mg or placebo in combination with a platinum

(either of CDDP 75 mg/m2 or CBDCA [AUC 5 mg·mL/min]) + PEM 500 mg/m2 on Day 1 of each 3-week

cycle for 4 cycles, followed by an intravenous dose of pembrolizumab 200 mg or placebo and PEM 500 mg/m2

intravenously every 3 weeks. The treatment was continued for a total of 35 cycles, or until disease progression

was observed or a withdrawal criterion was met.


including 4 and 6 Japanese patients, respectively) were included in the intention-to-treat (ITT) population and

used for the efficacy analyses. Of these 616 patients, 9 (5 in the pembrolizumab group and 4 in the placebo

group) did not receive the study drug, and the remaining 607 (405 in the pembrolizumab group and 202 in the

placebo group, including 4 and 6 Japanese patients, respectively) were included in the safety analysis set.

When the study was planned, the single primary endpoint was progression-free survival (PFS) as assessed

centrally using the Response Evaluation Criteria in Solid Carcinomas (RECIST) Ver. 1.1, and an interim PFS

analysis was planned when 50% of the target number of PFS events (373 events) had occurred. However, since

************* was observed in *************************** of ********************, the study

protocol was amended to add OS, which had been set as a secondary endpoint when the study was planned, as

another primary endpoint, and to conduct the interim PFS analysis and the first interim OS analysis when a

total of 370 PFS events have occurred, the final PFS analysis and the second OS interim analysis when a total

of 468 PFS events have occurred, and the final OS analysis when a total of 416 OS events have occurred

(Protocol Amendment Version 7 [dated November 6, 2017]).

In the PFS and OS analyses, the prespecified overall one-sided α level was allocated to the composite primary

endpoints (i.e., either 0.95% or 2.5% for PFS, and either 1.55% or 2.5% for OS), based on the results of the

PFS or OS analysis, according to the graphical approach of Maurer and Bretz (Stat Biopharm Res 2013;5:311-

20). A Lan-DeMets α spending function of the O’Brien-Fleming type was used to control the probability of a

type I error associated with the interim analysis.

The first interim analysis of OS, one of the primary efficacy endpoints, was performed (data cutoff date:

November 8, 2017). The results of the first interim OS analysis (Table 8) and the Kaplan-Meier curves of OS

(Figure 6) are shown below.

26


Table 8. First interim OS analysis (ITT population [data cutoff date: November 8, 2017])

Pembrolizumab Placebo N 410 206

Number of events (%) 127 (31.0) 108 (52.4) Median [95% CI] (months) — [—, —] 11.3 [8.7, 15.1]

Hazard ratio [95% CI]*1 0.49 [0.38, 0.64] P value (one-sided)*2 <0.00001

—, Not reached; *1, Cox regression stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a platinum (CDDP, CBDCA), and smoking history (yes, no); *2, Log-rank test stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a platinum (CDDP, CBDCA), and smoking history (yes, no), with a one-sided significance level of 0.00128

Figure 6. Kaplan-Meier curves for the first interim OS analysis

(ITT population [data cutoff date: November 8, 2017])

Table 9 shows the results of the interim analysis of PFS, another primary endpoint (data cutoff date: November

8, 2017).

Pembrolizumab

Placebo

Pembrolizumab

Placebo

27


Table 9. Interim PFS analysis (RECIST Ver. 1.1, central assessment, ITT population [data cutoff date: November 8, 2017])

Pembrolizumab Placebo

N 410 206 Number of events (%) 244 (59.5) 166 (80.6)

Median [95% CI] (months) 8.8 [7.6, 9.2] 4.9 [4.7, 5.5] Hazard ratio [95% CI]*1 0.52 [0.43, 0.64]

P value (one-sided)*2 <0.00001 —, Not reached; *1, Cox regression stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a platinum (CDDP, CBDCA), and smoking history (yes, no); *2, Log-rank test stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a platinum (CDDP, CBDCA), and smoking history (yes, no), with a one-sided significance level of 0.00559

The safety analysis indicated that 27 of 405 patients (6.7%) in the pembrolizumab group and 12 of 202 patients

(5.9%) in the placebo group died during the treatment period or within 90 days after the last dose (no Japanese

patients died). The causes of death in the pembrolizumab group were death and pneumonitis in 3 patients each,

intestinal ischaemia in 2 patients, and cardiac arrest/respiratory failure, cardiac arrest, acute kidney

injury/neutropenic sepsis, acute kidney injury, neutropenic sepsis, cardiac failure, cardiopulmonary failure,

cerebral infarction, chronic obstructive pulmonary disease, encephalopathy, haemoptysis, ischaemic stroke,

lung infection, mesenteric artery embolism, myocardial infarction, peritonitis, pneumocystis jirovecii

pneumonia, pneumonia, and septic shock in 1 patient each. The causes of death in the placebo group were

pneumonia/respiratory failure, hypokalaemia/supraventricular tachycardia, pneumonia, respiratory failure,

cerebral haemorrhage, death, disseminated intravascular coagulation, haemoptysis, haemorrhage intracranial,

multiple organ dysfunction syndrome, renal failure, and septic shock in 1 patient each. A causal relationship to

study drug could not be ruled out for the pneumonitis in 3 patients, acute kidney injury in 2 patients, and death,

encephalopathy, neutropenic sepsis, and pneumonia in 1 patient each in the pembrolizumab group, and the

pneumonia and septic shock in 1 patient each in the placebo group.

7.2.1.2.2 Global phase III study (CTD 5.3.5.1.4 and 5.3.5.1.5, Study 407, ongoing since August 2016

[data cutoff date: April 3, 2018])

A double-blind, randomized, comparative study was conducted at 125 sites in 17 countries and regions,

including Japan, to evaluate the efficacy and safety of pembrolizumab versus placebo, in combination with

platinum-based chemotherapy (CBDCA/PTX or CBDCA/nab-PTX) in chemotherapy-naïve patients with

unresectable, advanced or recurrent SQ-NSCLC (target sample size, 560 subjects).

Patients received an intravenous dose of pembrolizumab 200 mg or placebo in combination with platinum-

based chemotherapy (CBDCA [AUC 6 mg·mL/min] and either of PTX 200 mg/m2 or nab-PTX 100 mg/m2)

on Day 1 of each 3-week cycle for 4 cycles, followed by an intravenous dose of pembrolizumab 200 mg or

placebo every 3 weeks. The treatment was continued for a total of 35 cycles, or until disease progression was

observed or a withdrawal criterion was met.


including 22 and 28 Japanese patients, respectively) were included in the ITT population and used for the

efficacy analyses. Of the 559 patients, 1 patient in the placebo group did not receive the study drug, and the

28


remaining 558 (278 in the pembrolizumab group and 280 in the placebo group, including 22 and 28 Japanese

patients, respectively) were included in the safety analysis set.

When the study was planned, the composite primary endpoints were PFS as assessed centrally using RECIST

Ver. 1.1, and OS, while the response rate was a secondary endpoint. An interim analysis for efficacy was

planned. The final response rate analysis was to be conducted when the 200th enrolled and randomized patient

had been followed up for 28 weeks, the final PFS analysis and the interim OS analysis when a total of 331 PFS

events had occurred, and the final OS analysis when a total of 334 OS events had occurred. However, since

************* was observed in *************************** of ********************, the study

protocol was amended to change the prespecified allocation of the significance level for PFS and OS analyses,

and to conduct the final response rate analysis when the 200th enrolled and randomized patient has been

followed up for 28 weeks, the interim PFS analysis and the first interim OS analysis when a total of 332 PFS

events have occurred, the final PFS analysis and the second interim OS analysis when a total of 415 PFS events

have occurred, and the final OS analysis when a total of 361 OS events have occurred (Protocol Amendment

Version 2 [dated October 26, 2017]).

In the PFS and OS analyses, the prespecified overall one-sided α level was allocated to the composite primary

endpoints (i.e., 0.5% or 2.5% for the response rate, 1.0%, 1.5%, 2.0%, or 2.5% for PFS, and 1.0%, 2.0%, or

2.5% for OS) based on the results of the response rate, PFS, and OS analyses, according to the graphical

approach of Maurer and Bretz (Stat Biopharm Res 2013;5:311-20). A Lan-DeMets α spending function of the

O’Brien-Fleming type was used to control the probability of a type I error associated with the interim analysis.

The first interim analysis of OS, one of the primary efficacy endpoints, was performed (data cutoff date: April

3, 2018). The results of the first interim OS analysis (Table 10) and the Kaplan-Meier curves of OS (Figure 7)

are shown below.

Table 10. First interim OS analysis (ITT population [data cutoff date: April 3, 2018])

Pembrolizumab Placebo N 278 281

Number of events (%) 85 (30.6) 120 (42.7) Median [95% CI] (months) 15.9 [13.2, —] 11.3 [9.5, 14.8]

Hazard ratio [95% CI]*1 0.64 [0.49, 0.85] P value (one-sided)*2 0.0008

—, Not reached; *1, Cox regression stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a taxane-based chemotherapy (PTX, nab-PTX), and geographic region (East Asia, non-East Asia); *2, Log-rank test stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a taxane-based chemotherapy (PTX, nab-PTX), and geographic region (East Asia, non-East Asia), with a one-sided significance level of 0.0029

29


Figure 7. Kaplan-Meier curves for the first interim OS analysis

(ITT population [data cutoff date: April 3, 2018])

Table 11 shows the results of the final analysis of PFS, another primary endpoint (data cutoff date: April 3,

2018).

Table 11. Interim PFS analysis (RECIST Ver. 1.1, central assessment, ITT population [data cutoff date: April 3,

2018]) Pembrolizumab Placebo



P value (one-sided)*2 <0.0001 —, Not reached; *1, Cox regression stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a taxane-based chemotherapy (PTX, nab-PTX), and geographic region (East Asia, non-East Asia); *2, Log-rank test stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a taxane-based chemotherapy (PTX, nab-PTX), and geographic region (East Asia, non-East Asia), with a one-sided significance level of 0.008

Table 12 shows the results of the final analysis of the response rate, which was the secondary endpoint as

assessed centrally using RECIST Ver. 1.1, (data cutoff date: October 27, 2017).

Pembrolizumab

Placebo

Placebo

Pembrolizumab

30


Table 12. Best overall response and response rate (RECIST Ver. 1.1, central assessment, ITT population [data cutoff date: October 27, 2017])

Best overall responses n (%)


Placebo N = 103

CR 0 (0) 2 (1.9) PR 59 (58.4) 34 (33.0) SD 26 (25.7) 36 (35.0) PD 7 (6.9) 16 (15.5) NE 9 (9.0) 15 (14.6)

Response (CR + PR) (Response rate [95% CI] [%]) 59 (58.4 [48.2, 68.1]) 36 (35.0 [25.8, 45.0]) Difference estimator [95% CI]* 23.6 [9.9, 36.4]

P value (one-sided)* 0.0004 *, Miettinen and Nurminen test, stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a taxane-based chemotherapy (PTX, nab-PTX), and geographic region (East Asia, non-East Asia), with a one-sided significance level of 0.005

The safety analysis indicated that 23 of 278 patients (8.3%, including 1 of 22 Japanese patients [4.5%]) in the

pembrolizumab group, and 18 of 280 patients in the placebo group (6.4%, including 1 of 28 Japanese patients

[3.6%]) died during the treatment period or within 90 days after the last dose. The causes of death in the

pembrolizumab group were death in 4 patients, respiratory failure and sepsis in 3 patients each, cardiac arrest

and pulmonary haemorrhage in 2 patients each, and cardiac failure, circulatory collapse, hepatic failure,

intestinal perforation, lung abscess, necrotising fasciitis, pneumonia, pneumonitis, and pulmonary sepsis in 1

patient each. The causes of death in the placebo group were death and septic shock in 3 patients each, cardio-

respiratory arrest in 2 patients, and acute kidney injury, cardiac arrest, haemothorax, multiple organ dysfunction

syndrome, pleural effusion, pneumonia, pneumonitis, pulmonary haemorrhage, pulmonary mycosis, and sepsis

in 1 patient each. A causal relationship to the study drug could not be ruled out for the sepsis in 3 patients,

death in 2 patients, and hepatic failure, necrotising fasciitis, pneumonitis, pulmonary haemorrhage, and

respiratory failure in 1 patient each in the pembrolizumab group, and the septic shock in 2 patients, and acute

kidney injury, multiple organ dysfunction syndrome, pulmonary haemorrhage, and pneumonia in 1 patient each

in the placebo group (The causes of death in 2 Japanese patients were pneumonitis in 1 patient in the

pembrolizumab group and pulmonary haemorrhage in 1 patient in the placebo group, and a causal relationship

to the study drug could not be ruled out for both the events.)

7.2.1.2.3 Global phase III study (CTD 5.3.5.1.6, Study 042, ongoing since ** 20** [data cutoff date:

February 26, 2018])

An open-label, randomized, comparative study was conducted at 196 sites in 32 countries and regions,

including Japan, to evaluate the efficacy and safety of pembrolizumab versus the standard chemotherapy, in

chemotherapy-naïve patients with unresectable, advanced or recurrent PD-L1 positive (TPS ≥1%) NSCLC

(target sample size, 1240 patients).

Patients in the pembrolizumab group received an intravenous dose of pembrolizumab 200 mg on Day 1 of each

3-week cycle for up to 35 cycles. Patients in the chemotherapy group received CBDCA (AUC 5 or 6

mg·mL/min) and either of PEM 500 mg/m2 or PTX 200 mg/m2 intravenously for 6 cycles, followed by PEM

31


500 mg/m2 intravenously.8) The treatment was continued until disease progression was observed or a

withdrawal criterion was met.

All 1274 enrolled patients (637 in the pembrolizumab group and 637 in the chemotherapy group, including 47

and 46 Japanese patients, respectively) were included in the ITT population and used for the efficacy analyses.

Of these 1274 patients, 23 patients (1 in the pembrolizumab group and 22 in the chemotherapy group) did not

receive the study drug, and the remaining 1251 patients (636 in the pembrolizumab group and 615 in the

chemotherapy group, including 47 and 44 Japanese patients, respectively) were included in the safety analysis

set.

When the study was planned (dated ** **, 20**), the primary endpoint was OS in PD-L1 positive (TPS ≥50%)

patients, and both an interim analysis for futility in PD-L1 positive (TPS ≥1% and <50%) patients and 2 interim

analyses for efficacy were planned. (For protocol amendments, as well as the types and timing of the analyses,

see the table below.) However, since **********************************************************

*********************** was observed in ***************************************************

of **********, the study protocol was amended to cancel the interim analysis for futility in PD-L1 positive

(TPS ≥1% and <50%) patients (Protocol Amendment Version 2 [dated ** **, 20**]), to add a primary analysis

set comprising PD-L1 positive (TPS ≥20%) patients, and to conduct a single interim analysis for efficacy when

a total of 250 OS events have occurred in patients with PD-L1 positive (TPS ≥50%), and the last enrolled and

randomized patient has been followed up for 6 months (Protocol Amendment Version 3 [dated ** **, 20**]).

Furthermore, since *************************** *************** ******************

****************** ****************** **************** for ***** *****, the study protocol was

amended to change the timing of the final OS analysis from the time when a total of 340 OS events have

occurred in patients with PD-L1 positive (TPS ≥50%) to the time when the first enrolled and randomized

patient has been followed up for 45 months, and to conduct the second interim analysis for efficacy when the

first enrolled and randomized patient has been followed up for 38 months (Protocol Amendment Version 6

[dated ** **, 20**]).

8) In patients with NSQ-NSCLC who were categorized as complete response (CR), partial response (PR), or stable disease (SD) by

imaging-defined tumor assessment, according to the modified RECIST Ver. 1.1 performed after at least 4 cycles of CBDCA + PEM or CBDCA + PTX, PEM was administered thereafter.

32


Protocol Type of analysis Timing of analysis

Original version

Interim analysis (First)

When a total of 75 OS events have occurred in PD-L1 positive (TPS ≥1% and <50%) patients

Interim analysis (Second)

When (a) a total of 315 PFS events have occurred in PD-L1 positive (TPS ≥50%) patients, (b) a total of 100 OS events have occurred in PD-L1 positive (TPS ≥50%) patients, and (c) patient enrollment has been completed

Interim analysis (Third)

When a total of 283 OS events have occurred in PD-L1 positive (TPS ≥50%) patients

Final analysis When a total of 354 OS event have occurred in PD-L1 positive (TPS ≥50%) patients

Amendment Version 2 [dated

** **, 20**]


When a total of 187 OS events have occurred in PD-L1 positive (TPS ≥50%) patients and patient enrollment has been completed


When a total of 272 OS events have occurred in PD-L1 positive (TPS ≥50%) patients

Final analysis When a total of 340 OS events have occurred in PD-L1 positive (TPS ≥50%) patients


** **, 20**]

Interim analysis When a total of 250 OS events have occurred in PD-L1 positive (TPS ≥50%) patients, and the last enrolled and randomized patient has been followed up for 6 months

Final analysis When a total of 340 OS events have occurred in PD-L1 positive (TPS ≥50%) patients


** **, 20**]


When a total of 250 OS events have occurred in PD-L1 positive (TPS ≥50%) patients, and the last enrolled and randomized patient has been followed up for 6 months


When the first enrolled and randomized patient has been followed up for 38 months

Final analysis When the first enrolled and randomized patient has been followed up for 45 months

To adjust for multiplicity concerns arising from 3 analysis populations, the analysis was performed sequentially,

first in PD-L1 positive patients with TPS ≥50%, then in PD-L1 positive patients with TPS ≥20%, and finally

in PD-L1 positive patients with TPS ≥1%. The type I error associated with the interim analyses was controlled

through the use of a Lan-DeMets α spending function of the O’Brien-Fleming type for the first interim analysis,

and through the use of a Hwang-Shih-DeCani α spending function (γ parameter, −0.90239)) for the second

interim analysis and the final analysis.

The results of the second interim OS analysis (Tables 13, 14, and 15) and the Kaplan-Meier curves of OS

(Figures 8, 9, and 10) in the primary analysis sets are shown below.

Table 13. Second interim OS analysis (PD-L1 positive [TPS ≥50%] population [data cutoff date: February 26,

2018]) Pembrolizumab Chemotherapy



P value (one-sided)*2 0.0003 *1, Cox regression stratified by ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia); *2, Log-rank test stratified by ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia), with a one-sided significance level of 0.01220

9) With the addition of the second interim analysis and the change in the timing of the final analysis, the parameter was adjusted so

that the α spent at the first interim analysis would not be affected.

33


Figure 8. Kaplan-Meier curves for the second interim OS analysis

(PD-L1 positive [TPS ≥50%] population [data cutoff date: February 26, 2018])





P value (one-sided)*2 0.0020 *1, Cox regression stratified by PD-L1 expression status (TPS ≥50%, ≥1% and <50%), ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia); *2, Log-rank test stratified by PD-L1 expression status (TPS ≥50%, ≥1% and <50%), ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia), with a one-sided significance level of 0.01198

Pembrolizumab

Chemotherapy

Chemotherapy

Pembrolizumab

34


Figure 9. Kaplan-Meier curves for the second interim OS analysis (PD-L1 positive [TPS ≥20%] population [data cutoff date: February 26, 2018])





P value (one-sided)*2 0.0018 *1, Cox regression stratified by PD-L1 expression status (TPS ≥50%, ≥1% and <50%), ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia); *2, Log-rank test stratified by PD-L1 expression status (TPS ≥50%, ≥1% and <50%), ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia), with a one-sided significance level of 0.01238

Pembrolizumab

Chemotherapy

Chemotherapy

Pembrolizumab

35




The safety analysis indicated that 70 of 636 patients (11.0%) in the pembrolizumab group and 46 of 615 patients

(7.5%, including 1 of 44 Japanese patients [2.3%]) in the chemotherapy group died during the treatment period

or within 90 days after the last dose. The causes of death in the pembrolizumab group were death in 10 patients,

pneumonia in 8 patients, pulmonary embolism in 6 patients, pulmonary haemorrhage in 4 patients, respiratory

failure in 3 patients, cardiac arrest, cardio-respiratory arrest, gastric ulcer haemorrhage, sepsis, and septic shock

in 2 patients each, and malignant neoplasm progression, accidental death, acute respiratory failure, brain injury,

cardiac failure, cardiac failure acute, cerebrovascular accident, chronic obstructive pulmonary disease, coma,

diverticulitis, embolism, encephalopathy, febrile neutropenia, haemoptysis, hypercalcaemia of malignancy,

hypovolaemic shock, ileus, intestinal ischaemia, ischaemic stroke, Klebsiella infection, lung infection,

myocardial infarction, peripheral artery occlusion, pleural effusion, pneumonitis, pulmonary artery thrombosis,

pulmonary sepsis, sudden death, and tumour embolism in 1 patient each. The causes of death in the

chemotherapy group were pneumonia in 7 patients, death and pulmonary embolism in 5 patients each,

respiratory failure in 3 patients, pulmonary haemorrhage, pulmonary oedema, and pulmonary sepsis in 2

patients each, abdominal sepsis, acute coronary syndrome, biliary sepsis, cardiac failure, cardio-respiratory

arrest, cardiopulmonary failure, cerebral infarction, cerebrovascular accident, completed suicide, dyspnoea,

haemoptysis, infection, ketoacidosis, neutropenic sepsis, pancytopenia, pleural effusion, respiratory distress,

septic shock, sinus tachycardia, and sudden cardiac death in 1 patient each. A causal relationship to the study

drug could not be ruled out for the death, pulmonary embolism, respiratory failure, sepsis, cardiac failure acute,

encephalopathy, haemoptysis, hypovolaemic shock, ileus, Klebsiella infection, malignant neoplasm

Pembrolizumab

Chemotherapy

Chemotherapy

Pembrolizumab

36


progression, pneumonitis, and sudden death in 1 patient each in the pembrolizumab group, and the pneumonia

in 4 patients, and cardiac failure, dyspnoea, infection, ketoacidosis, neutropenic sepsis, pancytopenia,

pulmonary embolism, pulmonary sepsis, respiratory distress, and septic shock in 1 patient each in the

chemotherapy group (A causal relationship to the study drug was denied for the death in the Japanese patient

in the chemotherapy group).

7.2.1.3 Foreign clinical studies

7.2.1.3.1 Foreign phase I/II study (CTD 5.3.5.1.2, 5.3.5.1.3, and 5.3.5.1.7, Cohorts A, C, and G of Study

021, ongoing since ** 20** [data cutoff date: November 7, 2016 for Cohort A, and ** **, 20**

for Cohorts C and G])

Cohorts A and C in Part I of Study 021 was an open-label, uncontrolled study conducted at 9 sites in 2 countries

and regions outside Japan to evaluate the safety, tolerability, and other aspects of pembrolizumab in

combination with CBDCA, etc. in chemotherapy-naïve patients with unresectable, advanced or recurrent

NSCLC (Cohort A; target sample size, 24 patients) or in chemotherapy-naïve patients with unresectable,

advanced or recurrent NSQ-NSCLC (Cohort C; target sample size, 24 patients). Cohort G in Part II of Study

021 was an open-label, randomized, comparative study conducted at 7 sites in the U.S. to evaluate the efficacy

and safety of pembrolizumab in combination with CBDCA + PEM versus CBDCA + PEM alone in

chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC (Cohort G; target sample

size, 108 patients).

In Part I, patients in Cohort A received an intravenous dose of pembrolizumab 2 or 10 mg/kg, in combination

with CBDCA (AUC 6 mg·mL/min) + PTX 175 or 200 mg/m2, on Day 1 of each 3-week cycle for 4 cycles,

followed by pembrolizumab 2 or 10 mg/kg intravenously every 3 weeks. Patients in Cohort C received an

intravenous dose of pembrolizumab 2 or 10 mg/kg, in combination with CBDCA (AUC 5 mg·mL/min) + PEM

500 mg/m2, on Day 1 of each 3-week cycle for 4 cycles, followed by pembrolizumab 2 or 10 mg/kg and PEM

500 mg/m2 intravenously every 3 weeks. In Part II, patients in Cohort G received CBDCA (AUC 5

mg·mL/min) + PEM 500 mg/m2 intravenously, with or without an intravenous dose of pembrolizumab 200 mg,

on Day 1 of each 3-week cycle for 4 cycles, followed by PEM 500 mg/m2 intravenously, with or without of an

intravenous dose of pembrolizumab 200 mg every 3 weeks. The treatment was continued for a total of 24

months, or until disease progression was observed or a withdrawal criterion was met.

In Part I, all 49 enrolled patients (Cohort A [13 in the 2 mg/kg group and 12 in the 10 mg/kg group], Cohort C

[12 in the 2 mg/kg group and 12 in the 10 mg/kg group]) received the study drug and were included in the

safety analysis set for Part I.

In Part II, all 123 patients enrolled in Cohort G (60 in the pembrolizumab + CBDCA + PEM group and 63 in

the CBDCA + PEM group) were included in the ITT population, and used for the efficacy analyses. Of these

123 patients, 2 patients (1 in the pembrolizumab + CBDCA + PEM group and 1 in the CBDCA + PEM group)

did not receive the study drug, and the remaining 121 (59 in the pembrolizumab + CBDCA + PEM group and

62 in the CBDCA + PEM group) were included in the safety analysis set.

37


The results for the response rate as assessed centrally using RECIST Ver. 1.1 (the primary endpoint in Part II)

are shown in Table 16.

Table 16. Best overall response and response rate in Cohort G of Study 021

(RECIST Ver. 1.1, central assessment, ITT population [data cutoff date: ** **, 20**])


Pembrolizumab + CBDCA + PEM N = 60

CBDCA + PEM N = 63

CR 1 (1.7) 1 (1.6) PR 33 (55.0) 19 (30.2) SD 19 (31.7) 24 (38.1) PD 2 (3.3) 11 (17.5) NE 5 (8.3) 8 (12.7)

Response (CR + PR) (Response rate [95% CI] [%]) 34 (56.7 [43.2, 69.4]) 20 (31.7 [20.6, 44.7]) Difference estimator [95% CI]* 24.8 [7.2, 40.9]

P value (one-sided) 0.0029 *, Miettinen and Nurminen test stratified by PD-L1 expression status (TPS ≥1%, <1%)

The safety analysis indicated that 5 of 170 patients (2.9%) (Cohort C, 2 of 24 patients [8.3%]; Cohort G, 3 of

121 patients [2.5%] [1 of 59 patients, 1.7% in the pembrolizumab + CBDCA + PEM group, 2 of 62 patients,

3.2% in the CBDCA + PEM group]) died during the treatment period or within 90 days after the last dose. The

causes of death, other than disease progression, were: death in 2 patients in Cohort C; and, sepsis in 1 patient

in the pembrolizumab + CBDCA + PEM group, and pancytopenia and sepsis in 1 patient each in the CBDCA

+ PEM group in Cohort G. A causal relationship to the study drug could not be ruled out for the sepsis in 1

patient in the pembrolizumab + CBDCA + PEM group, and the pancytopenia and sepsis in 1 patient each in

the CBDCA + PEM group in Cohort G.

7.2.R Outline of the review conducted by PMDA for the treatment of NSCLC

7.2.R.1 Efficacy

Among the evaluation data submitted by the applicant, PMDA considered that the following 3 global phase III

studies ((a) to (c)) were important, and decided that the efficacy and safety of pembrolizumab would be

evaluated primarily based on the results of these studies.

PMDA also decided that the efficacy of pembrolizumab in Japanese patients would be evaluated in terms of

the consistency between the overall study population and the Japanese subpopulation of each study, according

to “Basic Principles on Global Clinical Trials” (PFSB/ELD Notification No. 0928010 dated September 28,

2007) and “Basic Principles on Global Clinical Trials (Reference Cases)” (PFSB/ELD Administrative Notice,

dated September 5, 2012).

(a) Study 189 to evaluate the efficacy and safety of pembrolizumab versus placebo, in combination with a

platinum + PEM in chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC

(b) Study 407 to evaluate the efficacy and safety of pembrolizumab versus placebo, in combination with

CBDCA + PTX or CBDCA + nab-PTX in chemotherapy-naïve patients with unresectable, advanced or

recurrent SQ-NSCLC

38


(c) Study 042 to evaluate the efficacy and safety of pembrolizumab versus chemotherapy in chemotherapy-

naïve patients with unresectable, advanced or recurrent PD-L1 positive (TPS ≥1%) NSCLC

As a result of its review, summarized below, PMDA has concluded that the efficacy of pembrolizumab in the

treatment of NSCLC has been demonstrated in the target patient populations of (a) Study 189, (b) Study 407,

and (c) Study 042.

7.2.R.1.1 Selection of control group

The applicant's rationale for the control treatments selected in (a) Study 189, (b) Study 407, and (c) Study 042:

(a) When Study 189 was planned, the NCCN guidelines (v.3.2015), etc. recommended combination

chemotherapy of a platinum + PEM, followed by maintenance therapy with PEM for the treatment of the

target patient population of Study 189, based on reports that a platinum + PEM had been shown to be

highly effective in such patient populations (J Clin Oncol 2008;26:3543-51, J Clin Oncol 2013;31:2895-

902). A platinum + PEM (followed by maintenance therapy with PEM) was therefore selected as the

control treatment.

(b) When Study 407 was planned, the NCCN guidelines (v.7.2015), etc. recommended CBDCA + PTX or

CBDCA + nab-PTX for the treatment of the target patient population of Study 407, based on a report that

these chemotherapy regimens had been shown to be highly effective in such patient populations (J Clin

Oncol 2012;30:2055-62). CBDCA + PTX and CBDCA + nab-PTX were therefore selected as control

treatments.

(c) When Study 042 was planned, the NCCN guidelines (v.3.2014), etc. recommended CBDCA + PEM or

CBDCA + PTX, followed by maintenance therapy with PEM for the treatment of the target patient

population of Study 042, based on reports that these chemotherapy regimens had been shown to be highly

effective in such patient populations (e.g., N Engl J Med 2002;346:92-8, Lancet Oncol 2012;13:247-55).

CBDCA + PEM and CBDCA + PTX (followed by maintenance therapy with PEM) were therefore

selected as control treatments.


7.2.R.1.2 Efficacy endpoints and evaluation results

(a) Study 189:

Study 189 demonstrated the superiority of pembrolizumab to placebo in OS (one of the primary endpoints)

[see Section 7.2.1.2.1].

The results of the first interim OS analysis (Table 17) and the Kaplan-Meier curves of OS (Figure 11) in the

Japanese subpopulation of Study 189 are shown below.

39


Table 17. First interim OS analysis in Japanese patients (ITT population [data cutoff date: November 8, 2017]) Pembrolizumab Placebo


Median [95% CI] (months) — [4.1, —] — [7.6, —] Hazard ratio [95% CI]*1 0.85 [0.08, 9.44]

P value (one-sided)*2 0.46650 —, Not reached; *1, Non-stratified Cox regression; *2, Non-stratified log-rank test

Figure 11. Kaplan-Meier curves for the first interim OS analysis in Japanese patients

(ITT population [data cutoff date: November 8, 2017])

(b) Study 407:

Study 407 demonstrated the superiority of pembrolizumab to placebo in OS (one of the primary endpoints)

[see Section 7.2.1.2.2].

The results of the first interim OS analysis (Table 18) and the Kaplan-Meier curves of OS (Figure 12) in the

Japanese subpopulation of Study 407 are shown below.

Pembrolizumab

Placebo

Placebo

Pembrolizumab

40


Table 18. First interim OS analysis in Japanese patients (ITT population [data cutoff date: April 3, 2018]) Pembrolizumab Placebo


Median [95% CI] (months) — [—, —] — [7.5, —] Hazard ratio [95% CI]*1 0.33 [0.07, 1.66]


Figure 12. Kaplan-Meier curves for the first interim OS analysis in Japanese patients

(ITT population [data cutoff date: April 3, 2018])

(c) Study 042:

Study 042 demonstrated the superiority of pembrolizumab to chemotherapy in OS (the primary endpoint) [see

Section 7.2.1.2.3].

PMDA's view:

In Study 042, the addition of the second interim OS analysis, as well as changes in the timing of the final OS

analysis and the method to control the increased probability of a type I error associated with the conduct of

interim analyses were made after the first interim OS analysis. PMDA considered that the possible impacts of

these changes on the interpretation of analytical results could not be excluded and asked the applicant to explain

the results of the analysis performed based on the analytical plan before the changes (Protocol Amendment

Version 2).

Pembrolizumab

Placebo

Placebo

Pembrolizumab

41


The applicant's response:

Table 19 shows the significance levels computed in the second interim OS analysis, accounting for the amount

of α spent at the first interim analysis and using a Lan-DeMets α spending function of the O’Brien-Fleming

type, based on the analysis plan specified in Protocol Amendment Version 2. The statistically significant

superiority of pembrolizumab to the chemotherapy in OS was also found in the second interim analysis,

performed based on the statistical plan before the changes.

Table 19. Second interim OS analysis based on the analysis plan specified in Protocol Amendment Version 5

(data cutoff date: February 26, 2018)

Population Hazard ratio [95% CI] P value (one-

sided) Significance level

(one-sided) PD-L1 positive (TPS ≥50%) 0.69 [0.56, 0.85]*1 0.0003*3 0.00797 PD-L1 positive (TPS ≥20%) 0.77 [0.64, 0.92]*2 0.0020*4 0.00777 PD-L1 positive (TPS ≥1%) 0.81 [0.71, 0.93]*2 0.0018*4 0.00813

*1, Cox regression stratified by ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia); *2, Cox regression stratified by PD-L1 expression status (TPS ≥50%, ≥1% and <50%), ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia); *3, Log-rank test stratified by ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia); *4, Log-rank test stratified by PD-L1 expression status (TPS ≥50%, ≥1% and <50%), ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia)

The results of the second interim OS analysis (Tables 20, 21, and 22) and the Kaplan-Meier curves of OS

(Figures 13, 14, and 15) in the Japanese subpopulation of Study 042 are shown below.

Table 20. Second interim OS analysis in Japanese patients (PD-L1 positive [TPS ≥50%] population [data cutoff

date: February 26, 2018]) Pembrolizumab Chemotherapy


Median [95% CI] (months) 27.4 [13.4, —] 19.9 [5.0, —] Hazard ratio [95% CI]*1 0.87 [0.36, 2.14]


42








P value (one-sided)*2 0.5045

—, Not reached; *1, Non-stratified Cox regression; *2, Non-stratified log-rank test

Pembrolizumab

Chemotherapy

Chemotherapy

Pembrolizumab

43


Figure 14. Kaplan-Meier curves for the second interim OS analysis in Japanese patients (PD-L1 positive [TPS ≥20%] population [data cutoff date: February 26, 2018])






Pembrolizumab

Chemotherapy

Chemotherapy

Pembrolizumab

44


Figure 15. Kaplan-Meier curves for the second interim OS analysis in Japanese patients


PMDA's view:

PMDA has concluded that pembrolizumab has been demonstrated to be effective when administered in

combination with chemotherapy, to chemotherapy-naïve patients with unresectable, advanced or recurrent

NSCLC, and when administered as a monotherapy to chemotherapy-naïve patients with advanced or recurrent

PD-L1 positive (TPS ≥1%) NSCLC, in view of the following findings.

• In Study 189 in chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC and

in Study 407 in chemotherapy-naïve patients with unresectable, advanced or recurrent SQ-NSCLC,

pembrolizumab was superior to placebo in OS (one of the primary endpoints).

• In Study 042 in chemotherapy-naïve patients with unresectable, advanced or recurrent PD-L1 positive

(TPS ≥1%) NSCLC, pembrolizumab was superior to chemotherapy in OS (the primary endpoint).

• Although the sample size of the Japanese subpopulations in Studies 189, 407, and 042, and the numbers of

events occurring in these subpopulations were insufficient to appropriately evaluate the efficacy of

pembrolizumab in Japanese patients, based only on the results from the subpopulations, the results from

the Japanese subpopulations did not tend to clearly differ from those from the overall study populations.

Pembrolizumab

Chemotherapy

Chemotherapy

Pembrolizumab

45


7.2.R.2 Safety [For adverse events, see “7.4.2 Adverse events reported in clinical studies in patients

with NSCLC.”]

PMDA's view:

As a result of its review described in the section below, PMDA considers that the adverse events that require

special attention after the administration of pembrolizumab in combination with platinum-based chemotherapy

to chemotherapy-naïve patients with unresectable, advanced or recurrent NSCLC, or as a monotherapy to

chemotherapy-naïve patients with unresectable, advanced or recurrent PD-L1 positive (TPS ≥1%) NSCLC are

those identified as requiring attention at the regulatory reviews for the approved indications7) and should also

be closely monitored when pembrolizumab is administered.

PMDA's conclusion:

Although attention should be paid to the above events, pembrolizumab is also tolerable when administered in

combination with platinum-based chemotherapy to chemotherapy-naïve patients with unresectable, advanced

or recurrent NSCLC, or when administered as a monotherapy to chemotherapy-naïve patients with unresectable,

advanced or recurrent PD-L1 positive (TPS ≥1%) NSCLC, as long as they are followed up by physicians with

sufficient knowledge and experience in cancer chemotherapy, through monitoring of adverse events,

differential diagnosis and patient management in anticipation of adverse reactions due to an excessive immune

response, drug interruption, or other appropriate actions.

7.2.R.2.1 Differences in safety profile between Japanese and non-Japanese patients

The applicant's explanation about the safety of pembrolizumab, based on the safety data from Studies 189, 407,

and 042:

Table 23 shows a summary of the safety data from Studies 189, 407, and 042.

46


Table 23. Safety summary (Studies 189, 407, and 042)

n (%) Study 189*1 Study 407*1 Study 042*2


Placebo N = 202


Placebo N = 280


Chemotherapy N = 615

All adverse events 404 (99.8) 200 (99.0) 273 (98.2) 274 (97.9) 610 (95.9) 606 (98.5) Grade ≥3 adverse events 272 (67.2) 133 (65.8) 194 (69.8) 191 (68.2) 318 (50.0) 351 (57.1) Adverse events leading to death

27 (6.7) 12 (5.9) 23 (8.3) 18 (6.4) 70 (11.0) 46 (7.5)

Serious adverse events 202 (49.9) 95 (47.0) 113 (40.6) 107 (38.2) 259 (40.7) 187 (30.4) Adverse events leading to drug discontinuation

Pembrolizumab or placebo

82 (20.2) 21 (10.4) 48 (17.3) 22 (7.9) 122 (19.2) —

Chemotherapy 96 (23.7) 26 (12.9) 44 (15.8) 29 (10.4) — 89 (14.5) Adverse events leading to drug interruption

Pembrolizumab or placebo

213 (52.6) 92 (45.5) 141 (50.7) 107 (38.2) 212 (33.3) —

Chemotherapy 199 (49.1) 91 (45.0) 132 (47.5) 131 (46.8) — 225 (36.6) Adverse events leading to dose reduction

Chemotherapy 68 (16.8) 20 (9.9) 63 (22.7) 49 (17.5) — 64 (10.4) *1, Pembrolizumab or placebo was administered in combination with platinum-based chemotherapy; *2, Pembrolizumab monotherapy or platinum-based chemotherapy


than in the placebo group were rash (20.2% [82 of 405 patients] in the pembrolizumab group, 11.4% [23 of

202 patients] in the placebo group), oedema peripheral (19.3% [78 of 405 patients], 12.9% [26 of 202 patients]),

and lacrimation increased (17.0% [69 of 405 patients], 10.9% [22 of 202 patients]). Grade ≥3 adverse events

reported with a ≥2% higher incidence in the pembrolizumab group than in the placebo group were neutropenia

(15.8% [64 of 405 patients], 11.9% [24 of 202 patients]), febrile neutropenia (6.7% [27 of 405 patients], 2.0%

[4 of 202 patients]), asthenia (6.2% [25 of 405 patients], 3.5% [7 of 202 patients]), fatigue (5.7% [23 of 405

patients], 2.5% [5 of 202 patients]), diarrhoea (5.2% [21 of 405 patients], 3.0% [6 of 202 patients]), and acute

kidney injury (2.0% [8 of 405 patients], 0%). Serious adverse events with a ≥2% higher incidence in the

pembrolizumab group than in the placebo group were febrile neutropenia (5.7% [23 of 405 patients], 2.0% [4

of 202 patients]), neutropenia (2.5% [10 of 405 patients], 0.5% [1 of 202 patients]), and acute kidney injury

(2.0% [8 of 405 patients], 0%). The adverse event that led to drug discontinuation with a ≥2% higher incidence

in the pembrolizumab group than in the placebo group was acute kidney injury (2.0% [8 of 405 patients], 0%).

Adverse events that led to drug interruption with a ≥2% higher incidence in the pembrolizumab group than in

the placebo group were thrombocytopenia (4.7% [19 of 405 patients], 1.0% [2 of 202 patients]), diarrhoea

(4.0% [16 of 405 patients], 1.0% [2 of 202 patients]), fatigue (3.5% [14 of 405 patients], 1.5% [3 of 202

patients]), and febrile neutropenia (2.2% [9 of 405 patients], 0%). No adverse events leading to death with a

2% higher incidence in the pembrolizumab group than in the placebo group.


than in the placebo group were alopecia (46.0% [128 of 278 patients] in the pembrolizumab group, 36.4% [102

of 280 patients] in the placebo group), thrombocytopenia (30.6% [85 of 278 patients], 23.2% [65 of 280

patients]), diarrhoea (29.9% [83 of 278 patients], 23.2% [65 of 280 patients]), arthralgia (20.5% [57 of 278

47


patients], 14.3% [40 of 280 patients]), pruritus (12.9% [36 of 278 patients], 7.5% [21 of 280 patients]), and

taste abnormality (9.0% [25 of 278 patients], 3.9% [11 of 280 patients]). The Grade ≥3 adverse events reported

with a ≥2% higher incidence in the pembrolizumab group than in the placebo group was pneumonitis (2.5% [7

of 278 patients], 0.4% [1 of 280 patients]). Adverse events that led to drug interruption with a ≥2% higher

incidence in the pembrolizumab group than in the placebo group were thrombocytopenia (14.4% [40 of 278

patients], 8.9% [25 of 280 patients]), neutropenia (11.5% [32 of 278 patients], 8.6% [24 of 280 patients]), and

pneumonitis (2.2% [6 of 278 patients], 0 %). No serious adverse events, adverse events leading to drug

discontinuation or death were reported with a ≥2% higher incidence in the pembrolizumab group than in the

placebo group.


than in the chemotherapy group were dyspnoea (16.5% [105 of 636 patients] in the pembrolizumab group,

11.4% [70 of 615 patients] in the chemotherapy group), cough (15.6% [99 of 636 patients], 10.6% [65 of 615

patients]), hypothyroidism (12.1% [77 of 636 patients], 1.5% [9 of 615 patients]), hyperthyroidism (6.1% [39

of 636 patients]), 0.7% [4 of 615 patients]), and pneumonitis (7.4% [47 of 636 patients]), 0.2% [1 of 615

patients]). The Grade ≥3 adverse events reported with a ≥2% higher incidence in the pembrolizumab group

than in the chemotherapy group was pneumonitis (3.1% [20 of 636 patients]), 0%). Serious adverse events

reported with a ≥2% higher incidence in the pembrolizumab group than in the chemotherapy group were

pneumonia (7.4% [47 of 636 patients], 5.2% [32 of 615 patients]) and pneumonitis (3.9% [25 of 636 patients],

0.2% [1 of 615 patients]. The adverse event that led to drug discontinuation with a ≥2% higher incidence in

the pembrolizumab group than in the chemotherapy group was pneumonitis (3.0% [19 of 636 patients], 0%).

Adverse events that led to drug interruption with a ≥2% higher incidence in the pembrolizumab group than in

the chemotherapy group were pneumonitis (3.1% [20 of 636 patients], 0.2% [1 of 615 patients] and

hypothyroidism (2.2% [14 of 636 patients], 0%). No adverse events led to death with a ≥2% higher incidence

in the pembrolizumab group than in the chemotherapy group.

The applicant's explanation about the differences in the safety profile of pembrolizumab between (a)

chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC, (b) chemotherapy-naïve

patients with unresectable, advanced or recurrent SQ-NSCLC, or (c) chemotherapy-naïve patients with

unresectable, advanced or recurrent PD-L1 positive (TPS ≥1%) NSCLC, and patients treated for the approved

indications was as follows:

The results of the comparisons of the incidences of adverse events between patients in the pembrolizumab

group of Study 189, 407, or 042 and patients who were treated with pembrolizumab for the approved

indications are shown in Table 6 [see Section 7.1.R.2.1].

(a) Study 189:

Adverse events of any grade reported with a ≥10% higher incidence in chemotherapy-naïve patients with

unresected, advanced or recurrent NSQ-NSCLC than in any of the other patient populations, treated for the

approved indications, were nausea (chemotherapy-naïve, unresectable, advanced or recurrent NSQ-NSCLC,

48


55.6%; PD-L1 positive NSCLC, 20.2%, unresectable malignant melanoma, 25.3%; cHL, 13.3%; radically

unresectable urothelial carcinoma, 20.7%), anaemia (46.2%, 10.3%, 12.5%, 9.0%, 17.3%), constipation

(34.8%, 16.4%, 19.2%, 9.5%, 18.8%), neutropenia (27.2%, 0.6%, 0.5%, 6.2%, 0%), thrombocytopenia (18.0%,

1.1%, 2.1%, 5.2%, 0.4%), and lacrimation increased (17.0%, 0.5%, 0.7%, 1.9%, 1.1%). Grade ≥3 adverse

events reported with a ≥5% higher incidence in chemotherapy-naïve patients with unresected, advanced or

recurrent NSQ-NSCLC than in any of the other patient populations were anaemia (16.3%, 2.8%, 3.7%, 3.8%,

8.3%), neutropenia (15.8%, 0%, 0.2%, 2.9%, 0%), thrombocytopenia (7.9%, 0.4%, 0.4%, 1.9%, 0%), febrile

neutropenia (6.7%, 0.2%, 0%, 1.0%, 0%), and asthenia (6.2%, 1.2%, 1.2%, 0%, 0.8%). The serious adverse

event reported with a ≥5% higher incidence in chemotherapy-naïve patients with unresected, advanced or

recurrent NSQ-NSCLC than in any of the other patient populations was febrile neutropenia (5.7%, 0.2%, 0%,

0%, 0%). Adverse events that led to drug (pembrolizumab or concomitant chemotherapy) interruption with a

≥5% higher incidence in chemotherapy-naïve patients with unresected, advanced or recurrent NSQ-NSCLC

than in any of the other patient populations were neutropenia (13.1%, 0%, 0.1%, 0%, 0%) and anaemia (7.4%,

0.8%, 0.8%, 0%, 0.4%). No adverse events led to death or drug (pembrolizumab or concomitant chemotherapy)

discontinuation, with a ≥5% higher incidence in chemotherapy-naïve patients with unresected, advanced or

recurrent NSQ-NSCLC than in any of the other patient populations.

The above comparisons revealed no clear differences in the safety profile of pembrolizumab between

chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC and patients treated for

the approved indications, although some adverse events and serious adverse events, etc. occurred more

frequently in chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC than in

patients treated for the approved indications, however, all of these adverse events were known adverse events

of CDDP, CBDCA, or PEM.

(b) Study 407:


unresected, advanced or recurrent SQ-NSCLC than in any of the other patient populations, treated for the

approved indications, were anaemia (chemotherapy-naïve, unresectable, advanced or recurrent SQ-NSCLC,

53.2%; PD-L1 positive NSCLC, 10.3%; unresectable malignant melanoma, 12.5%; cHL, 9.0%; radically

unresectable urothelial carcinoma, 17.3%), alopecia (46.0%, 1.3%, 2.1%, 2.4%, 0.8%), neutropenia (37.8%,

0.6%, 0.5%, 6.2%, 0%), nausea (35.6%, 20.2%, 25.3%, 13.3%, 20.7%), thrombocytopenia (30.6%, 1.1%, 2.1%,

5.2%, 0.4%), and neuropathy peripheral (20.5%, 1.9%, 1.8%, 3.8%, 0.4%). Grade ≥3 adverse events reported

with a ≥5% higher incidence in chemotherapy-naïve patients with unresected, advanced or recurrent SQ-

NSCLC than in any of the other patient populations were neutropenia (22.7%, 0%, 0.2%, 2.9%, 0%), anaemia

(15.5%, 2.8%, 3.7%, 3.8%, 8.3%), and neutrophil count decreased (6.1%, 0%, 0.1%, 0%, 0.4%). The serious

adverse event reported with a ≥5% higher incidence in chemotherapy-naïve patients with unresected, advanced

or recurrent SQ-NSCLC than in any of the other patient populations was febrile neutropenia (5.4%, 0.2%, 0%,

0%, 0%). Adverse events that led to drug (pembrolizumab or concomitant chemotherapy) interruption with a

≥5% higher incidence in chemotherapy-naïve patients with unresected, advanced or recurrent SQ-NSCLC than

in any of the other patient populations were anaemia (12.9%, 0.8%, 0.8%, 0%, 0.4%), neutropenia (19.4%, 0%,

49


0.1%, 0%, 0%), and thrombocytopenia (18.0%, 0%, 0.2%, 0%, 0%). No adverse events led to death or drug

(pembrolizumab or concomitant chemotherapy) discontinuation, with a ≥5% higher incidence in

chemotherapy-naïve patients with unresected, advanced or recurrent SQ-NSCLC than in any of the other

patient populations.

The above comparisons revealed no clear differences in the safety profile of pembrolizumab between

chemotherapy-naïve patients with unresectable, advanced or recurrent SQ-NSCLC and patients treated for the

approved indications, although some adverse events and serious adverse events, etc. occurred more frequently

in chemotherapy-naïve patients with unresectable, advanced or recurrent SQ-NSCLC than in patients treated

for the approved indications, however, all of these adverse events were known adverse events of CBDCA, PTX,

or nab-PTX.

(c) Study 042:


advanced or recurrent PD-L1 (TPS ≥1%) NSCLC than in any of the other patient populations, treated for the

approved indications, was pneumonia (chemotherapy-naïve, advanced or recurrent PD-L1 positive [TPS ≥1%]

NSCLC, 11.9%; PD-L1 positive NSCLC, 6.7%; unresectable malignant melanoma, 2.0%; cHL, 2.4%;

radically unresectable urothelial carcinoma, 4.5%). The Grade ≥3 adverse events reported with a ≥2% higher

incidence in chemotherapy-naïve patients with advanced or recurrent PD-L1 (TPS ≥1%) NSCLC than in any

of the other patient populations was pneumonia (7.4%, 4.7%, 0.8%, 0.5%, 2.6%). The serious adverse event

reported with a ≥2% higher incidence in chemotherapy-naïve patients with advanced or recurrent PD-L1 (TPS

≥1%) NSCLC than in any of the other patient populations was also pneumonia (7.4%, 4.7%, 1.1%, 1.4%, 3.4%).

No adverse events led to death, drug interruption or discontinuation, with a ≥2% higher incidence in

chemotherapy-naïve patients with advanced or recurrent PD-L1 (TPS ≥1%) NSCLC than in any of the other

patient populations.

The above comparisons revealed no clear differences in the frequently noted adverse events and serious adverse

events, etc. between chemotherapy-naïve patients with advanced or recurrent PD-L1 positive (TPS ≥1%)

NSCLC and patients treated for the approved indications, suggesting that the safety of pembrolizumab is

similar for these patient populations.

The applicant's explanation about the safety of pembrolizumab between Japanese and non-Japanese patients,

based on the safety data from patients receiving pembrolizumab in (a) Study 189, (b) Study 407, and (c) Study

042:

(a) Study 189:

Table 24 shows a summary of the safety data in Japanese patients (including the Japanese expanded cohort10))

and non-Japanese patients receiving pembrolizumab in Study 189.

10) In Part B of Study 011, 1 patient each died of ILD and pneumonitis, for which a causal relationship to the study drug could not be

ruled out. Accordingly, the enrollment of Japanese patents in Study 189 was suspended, and an expanded cohort composed of

50


Table 24. Safety summary (Study 189) n (%)

Japanese N = 25

Non-Japanese N = 401

All adverse events 25 (100) 400 (99.8) Grade ≥3 adverse events 13 (52.0) 270 (67.3) Adverse events leading to death 0 27 (6.7) Serious adverse events 5 (20.0) 200 (49.9) Adverse events leading to drug discontinuation

Pembrolizumab 3 (12.0) 81 (20.2) Chemotherapy 3 (12.0) 96 (23.9)

Adverse events leading to drug interruption Pembrolizumab 16 (64.0) 211 (52.6) Chemotherapy 15 (60.0) 198 (49.4)

Adverse events leading to dose reduction Pembrolizumab — — Chemotherapy 3 (12.0) 67 (16.7)

In the pembrolizumab group of Study 189, adverse events of any grade reported with a ≥10% higher incidence

in Japanese patients than in non-Japanese patients were nausea (Japanese, 72.0% [18 of 25 patients]; 55.6%

[223 of 401 patients]), constipation (64.0% [16 of 25 patients], 34.4% [138 of 401 patients]), decreased appetite

(56.0% [14 of 25 patients]), 27.9% [112 of 401 patients]), white blood cell count decreased (40.0% [10 of 25

patients], 6.2% [25 of 401 patients]), ALT increased (28.0% [7 of 25 patients], 11.5% [46 of 401 patients]),

neutrophil count decreased (28.0% [7 of 25 patients], 2.7% [11 of 401 patients]), lymphocyte count decreased

(28.0% [7 of 25 patients], 2.5% [10 of 401 patients]), hiccups (28.0% [7 of 25 patients], 4.7% [19 of 401

patients]), AST increased (24.0% [6 of 25 patients], 8.7% [35 of 401 patients]), stomatitis (24.0% [6 of 25

patients], 8.2% [33 of 401 patients]), dry skin (24.0% [6 of 25 patients], 4.2% [17 of 401 patients]), malaise

(24.0% [6 of 25 patients], 0.5% [2 of 401 patients]), insomnia (20.0% [5 of 25 patients], 5.7% [23 of 401

patients]), alopecia (16.0% [4 of 25 patients], 5.0% [20 of 401 patients]), hepatic function abnormal (12.0% [3

of 25 patients], 0.2% [1 of 401 patients]), and diabetes mellitus (12.0% [3 of 25 patients], 0.2% [1 of 401

patients]). Grade ≥3 adverse events reported with a ≥5% higher incidence in Japanese patients than in non-

Japanese patients were neutrophil count decreased (20.0% [5 of 25 patients], 1.7% [7 of 401 patients]),

lymphocyte count decreased (20.0% [5 of 25 patients], 0.5% [2 of 401 patients]), white blood cell count

decreased (16.0% [4 of 25 patients]), 1.5% [6 of 401 patients]), and hyponatraemia (12.0% [3 of 25 patients]),

1.7% [7 of 401 patients]). Adverse events that led to drug interruption with a ≥5% higher incidence in Japanese

patients than in non-Japanese patients were white blood cell count decreased (20.0% [5 of 25 patients]), 1.7%

[7 of 401 patients]), neutrophil count decreased (16.0% [4 of 25 patients], 0.5% [2 of 401 patients]), and

lymphocyte count decreased (12.0% [3 of 25 patients], 0%). No adverse events leading to death, serious adverse

events, or adverse events leading to drug discontinuation were reported, with a ≥5% higher incidence in

Japanese patients than in non-Japanese patients.

Japanese patients was to be added to enroll 40 Japanese patients (the target sample size for Japanese patients when Study 189 was planned), including the 10 Japanese patients who had been enrolled (Protocol Amendment Version 6 [dated ** **, 20**]).

51


In the placebo group of Study 189, adverse events of any grade reported in ≥2 Japanese patients and with a

≥10% higher incidence in Japanese patients than in non-Japanese patients were constipation (73.3% [11

patients], 30.6% [60 patients]), anaemia (66.7% [10 patients], 45.4% [89 patients]), decreased appetite (60.0%

[9 patients], 29.1% [57 patients]), leukopenia (46.7% [7 patients], 5.1% [10 patients]), malaise (33.3% [5

patients], 1.5% [3 patients]), ALT increased (33.3% [5 patients], 8.2% [16 patients]), AST increased (33.3%

[5 patients], 5.6% [11 patients]), taste abnormality (33.3% [5 patients], 9.2% [18 patients]), hiccups (33.3% [5

patients], 1.5% [3 patients]), thrombocytopenia (26.7% [4 patients], 13.8% [27 patients]), stomatitis (26.7% [4

patients], 7.7% [15 patients]), oedema peripheral (26.7% [4 patients], 12.2% [24 patients]), white blood cell

count decreased (20.0% [3 patients], 5.6% [11 patients]), anxiety (20.0% [3 patients], 4.6% [9 patients]),

neutrophil count decreased (13.3% [2 patients], 2.0% [4 patients]), platelet count decreased (13.3% [2 patients],

0.5% [1 patient]), tumour pain (13.3% [2 patients], 1.0% [2 patients]), peripheral sensory neuropathy (13.3%

[2 patients]), 1.0% [2 patients]), and ILD (13.3% [2 patients], 0%). Grade ≥3 adverse events reported in ≥2

Japanese patients were anaemia (20.0% [3 patients], 14.3% [28 patients]) and neutropenia (13.3% [2 patients],

11.2% [22 patients]). The serious adverse event reported in ≥2 Japanese patients was ILD (13.3% [2 patients],

0%). No Japanese patient died due to adverse events.

(b) Study 407:

Table 25 shows a summary of the safety data from Japanese and non-Japanese patients receiving

pembrolizumab in Study 407.


n (%)

Japanese N = 22


All adverse events 22 (100) 251 (98.0) Grade ≥3 adverse events 19 (86.4) 175 (68.4) Adverse events leading to death 1 (4.5) 22 (8.6) Serious adverse events 13 (59.1) 100 (39.1) Adverse events leading to drug discontinuation

Pembrolizumab 7 (31.8) 41 (16.0) Chemotherapy 8 (36.4) 36 (14.1)

Adverse events leading to drug interruption Pembrolizumab 20 (90.9) 121 (47.3) Chemotherapy 15 (68.2) 117 (45.7)

Adverse events leading to dose reduction Chemotherapy 6 (27.3) 57 (22.3)


in Japanese patients than in non-Japanese patients were anaemia (Japanese, 86.4% [19 of 22 patients]; non-

Japanese, 50.4% [129 of 256 patients]), alopecia (77.3% [17 of 22 patients], 43.4% [111 of 256 patients]),

neutropenia (63.6% [14 of 22 patients], 35.5% [91 of 256 patients]), constipation (59.1% [13 of 22 patients],

19.9% [51 of 256 patients]), white blood cell count decreased (54.5% [12 of 22 patients], 7.4% [19 of 256

patients]), decreased appetite (50.0% [11 of 22 patients], 22.3% [57 of 256 patients]), malaise (40.9% [9 of 22

patients], 0.4% [1 of 256 patients]), platelet count decreased (40.9% [9 of 22 patients], 5.9% [15 of 256

patients]), peripheral sensory neuropathy (40.9% [9 of 22 patients], 9.0% [23 of 256 patients]), insomnia

52


(36.4% [8 of 22 patients], 7.8% [20 of 256 patients]), leukopenia (27.3% [6 of 22 patients], 7.0% [18 of 256

patients]), neutrophil count decreased (27.3% [6 of 22 patients], 7.0% [18 of 256 patients]), stomatitis (18.2%

[4 of 22 patients], 2.3% [6 of 256 patients]), hyponatraemia (18.2% [4 of 22 patients], 3.9% [10 of 256

patients]), hiccups (18.2% [4 of 22 patients], 3.5% [9 of 256 patients]), and rash maculo-papular (18.2% [4 of

22 patients], 1.6% [4 of 256 patients]). Grade ≥3 adverse events reported with a ≥5% higher incidence in

Japanese patients than in non-Japanese patients were neutropenia (40.9% [9 of 22 patients], 21.1% [54 of 256

patients]), anaemia (36.4% [8 of 22 patients], 13.7% [35 of 256 patients]), leukopenia (22.7% [5 of 22 patients],

3.1% [8 of 256 patients]), neutrophil count decreased (18.2% [4 of 22 patients], 5.1% [13 of 256 patients]),

white blood cell count decreased (18.2% [4 of 22 patients], 3.1% [8 of 256 patients]), hyponatraemia (18.2%

[4 of 22 patients], 2.3% [6 of 256 patients]), diarrhoea (9.1% [2 of 22 patients], 3.5% [9 of 256 patients]),

delirium (9.1% [2 of 22 patients], 0%), and pneumonitis(9.1% [2 of 22 patients], 2.0% [5 of 256 patients]).

The serious adverse event reported with a ≥5% higher incidence in Japanese patients than in non-Japanese

patients was pneumonitis (13.6% [3 of 22 patients], 1.6% [4 of 256 patients]). The adverse event that led to

drug discontinuation with a ≥5% higher incidence in Japanese patients than in non-Japanese patients was also

pneumonitis (9.1% [2 of 22 patients], 1.2% [3 of 256 patients]). Adverse events that led to drug interruption

with a ≥5% higher incidence in Japanese patients than in non-Japanese patients were anaemia (40.9% [9 of 22

patients], 5.9% [15 of 256 patients]), neutropenia (31.8% [7 of 22 patients], 9.8% [25 of 256 patients]),

leukopenia (18.2% [4 of 22 patients], 1.2% [3 of 256 patients]), platelet count decreased (18.2% [4 of 22

patients], 1.6% [4 of 256 patients]), white blood cell count decreased (18.2% [4 of 22 patients], 0.4% [1 of 256

patients]), diarrhoea (9.1% [2 of 22 patients], 1.2% [3 of 256 patients]), rash (9.1% [2 of 22 patients], 0.4% [1

of 256 patients]), and decreased appetite (9.1% [2 of 22 patients], 0%). No adverse events led to death with a

≥5% higher incidence in Japanese patients than in non-Japanese patients.

In the placebo group of Study 407, adverse events of any grade reported with a ≥10% higher incidence in

Japanese patients than in non-Japanese patients were anaemia (Japanese, 64.3% [18 patients]; non-Japanese,

50.4% [127 patients]), decreased appetite (64.3% [18 patients], 25.4% [64 patients]), alopecia (60.7% [17

patients], 33.7% [85 patients]), constipation (57.1% [16 patients], 17.9% [45 patients]), white blood cell count

decreased (46.4% [13 patients], 6.7% [17 patients]), peripheral sensory neuropathy (42.9% [12 patients], 9.5%

[24 patients]), neutrophil count decreased (32.1% [9 patients], 7.5% [19 patients]), malaise (32.1% [9 patients],

1.6% [4 patients]), arthralgia (28.6% [8 patients], 12.7% [32 patients]), pyrexia (25.0% [7 patients], 11.9% [30

patients]), platelet count decreased (17.9% [5 patients], 6.0% [15 patients]), stomatitis (14.3% [4 patients],

4.0% [10 patients]), GGT increased (14.3% [4 patients], 1.6% [4 patients]), and gingivitis (10.7% [3 patients],

0%). Grade ≥3 adverse events reported with a ≥5% higher incidence in Japanese patients than in non-Japanese

patients were neutropenia (25.0% [7 patients], 6.7% [17 patients]), white blood cell count decreased (17.9% [5

patients], 2.4% [6 patients]), and lung infection (7.1% [2 patients], 0.4% [1 patient]). The serious adverse event

reported with a ≥5% higher incidence in Japanese patients than in non-Japanese patients was lung infection

(7.1% [2 patients], 0%). No adverse events led to death with a ≥5% higher incidence in Japanese patients than

in non-Japanese patients.

53


In the pembrolizumab group, the incidence of ILD-like events11) was higher in Japanese patients (18.2% [4 of

22 patients]) than in non-Japanese patients (5.5% [14 of 256 patients]). However, the incidence of ILD-like

events was also higher in the placebo group in Japanese patients (10.7% [3 of 28 patients] than in non-Japanese

patients [1.2% [3 of 252 patients]). Of the 4 Japanese patients experiencing an ILD-like event in the

pembrolizumab group, 1 died of the ILD-like event, for which a causal relationship to the study drug could not

be ruled out. However, a blinded assessment by the ILD evaluation committee found that the patient had

honeycomb lung at baseline, which was an exclusion criterion in Study 407, and that the ILD-like event that

led to death might have been associated with the honeycomb lung. The ILD-like events in the remaining 3

patients were not resolving (Grade 3), resolved (Grade 2), and improved (Grade 2) in 1 patient each.

The above results indicated that pembrolizumab in combination with chemotherapy may cause ILD-like events

more frequently than chemotherapy alone; therefore, attention should be paid to the risk of ILD-like events.

However, these ILD-like events are known adverse events of both pembrolizumab and chemotherapy.

Therefore, pembrolizumab in combination with platinum-based chemotherapy is also tolerable in Japanese

patients, as long as the occurrence of ILD-like events is carefully monitored and managed appropriately.

(c) Study 042:

Table 26 shows a summary of the safety data from Japanese and non-Japanese patients receiving

pembrolizumab in Study 042.


n (%) Japanese N = 47


All adverse events 41 (87.2) 569 (96.6) Grade ≥3 adverse events 19 (40.4) 299 (50.8) Adverse events leading to death 0 70 (11.9) Serious adverse events 17 (36.2) 242 (41.1) Adverse events leading to drug discontinuation 8 (17.0) 114 (19.4) Adverse events leading to drug interruption 13 (27.7) 199 (33.8)


in Japanese patients than in non-Japanese patients were constipation (Japanese, 23.4% [11 of 47 patients]; non-

Japanese, 11.2% [66 of 589 patients]), insomnia (23.4% [11 of 47 patients], 3.7% [22 of 589 patients]), rash

(23.4% [11 of 47 patients], 9.8% [58 of 589 patients]), malaise (19.1% [9 of 47 patients], 1.4% [8 of 589

patients]), back pain (19.1% [9 of 47 patients], 9.0% [53 of 589 patients]), pneumonitis (17.0% [8 of 47

patients], 6.6% [39 of 589 patients]), and renal impairment (10.6% [5 of 47 patients], 0.5% [3 of 589 patients]).

The Grade ≥3 adverse event reported with a ≥5% higher incidence in Japanese patients than in non-Japanese

patients was decreased appetite (6.4% [3 of 47 patients], 1.4% [8 of 589 patients]). The serious adverse event

reported with a ≥5% higher incidence in Japanese patients than in non-Japanese patients was pneumonitis

(10.6% [5 of 47 patients], 3.4% [20 of 589 patients]. The adverse event that led to drug interruption with a ≥5%

11) Events coded as MedDRA preferred terms of “acute interstitial pneumonitis,” “ILD,” “pneumonitis,” “idiopathic pneumonia

syndrome,” and “organising pneumonia” were counted.

54


higher incidence in Japanese patients than in non-Japanese patients was also pneumonitis (8.5% [4 of 47

patients], 2.7% [16 of 589 patients]). No adverse events led to death or drug discontinuation, with a ≥5% higher

incidence in Japanese patients than in non-Japanese patients.

In the pembrolizumab group, the incidence of ILD-like events11) was higher in Japanese patients (21.3% [10

of 47 patients]) than in non-Japanese patients (7.3% [43 of 589 patients]). Of the 10 Japanese patients

experiencing ILD-like events in the pembrolizumab group, 3 had Grade ≥3 events (Grade 3 in all of these

patients), and no patients died of ILD-like events. At the time of data cutoff, the ILD-like events occurring in

8 of the 10 patients, except those of Grade 1 and Grade 2 in 1 patient each, had improved or resolved. In view

of these results, pembrolizumab is tolerable in Japanese patients as well as non-Japanese patients, although

continuous attention should be paid to the risk of ILD-like events.

PMDA's view:

In Studies 189, 407, and 042, some adverse events were reported more frequently in the pembrolizumab group

than in (a) the control group or (b) patients treated for the approved indications. However, all of these events

were known adverse events of pembrolizumab or the concomitant chemotherapy. PMDA has therefore

concluded that pembrolizumab is tolerable in the following treatments of patients with NSCLC, as long as they

are followed up by physicians with sufficient knowledge and experience in cancer chemotherapy, through

monitoring of adverse events, differential diagnosis, and patient management in anticipation of adverse

reactions caused by an excessive immune response, drug interruption, or other appropriate actions. In Studies

189 and 407, Grade 3 or serious febrile neutropenia were more frequently reported in the pembrolizumab group

than in the placebo group. However, all of these events improved or resolved, and are known adverse events

of the concomitant chemotherapy. Therefore, the inclusion of any specific precautions in the package insert is

unnecessary, although information on the incidence of the adverse reactions should be provided to healthcare

professionals, through the package insert or other materials.


naïve patients with unresectable, advanced or recurrent NSCLC


advanced or recurrent PD-L1 positive (TPS ≥1%) NSCLC

All of the adverse events that were more frequently reported in Japanese patients than in non-Japanese patients

are known adverse events of pembrolizumab. In addition, considering the incidences of Grade ≥3 adverse

events, serious advance events, and other notable adverse events, PMDA has concluded that pembrolizumab

is tolerable in Japanese patients with unresectable, advanced or recurrent NSCLC, as long as appropriate

measures, such as drug interruption are taken.

7.2.R.3 Clinical positioning and indications

The present partial change application has been proposed to change the indication of pembrolizumab from the

approved indication of “unresectable, advanced or recurrent PD-L1 positive NSCLC” to “unresectable,

55


advanced or recurrent NSCLC,” and to include the following statements in the “Precautions for Indications”

section of the package insert.

• The efficacy and safety of pembrolizumab in adjuvant chemotherapy have not been established (unchanged

from the previous approval).

• The “Clinical Studies” section should be carefully read to understand the percentage of PD-L1-expressing

tumor cells (TPS). PD-L1 testing should be performed by a highly experienced pathologist or at a

laboratory facility using the approved in vitro diagnostic. When used as monotherapy, pembrolizumab

should be administered to patients whose tumors have been demonstrated to express PD-L1.


pembrolizumab, after carefully reading the “Clinical Studies” section to understand the status of EGFR

mutation or ALK fusion genes in patients enrolled in the clinical studies (unchanged from the previous

approval).

As a result of its review described in Sections “7.2 R.1 Efficacy,” “7.2.R.3 Safety,” and the section below,

PMDA concluded that the modified indication, "unresectable, advanced or recurrent NSCLC" proposed by the

applicant is appropriate and that the following precautionary statements should be included in the “Precautions

for Indications” section.



• When used as monotherapy, pembrolizumab should be administered to patients whose tumors have been

demonstrated to express PD-L1. The “Clinical Studies” section should be carefully read to understand the

percentage of PD-L1-expressing tumor cells (TPS). PD-L1 testing should be performed by a highly

experienced pathologist or at a laboratory facility using the approved in vitro diagnostic.




approval).

7.2.R.3.1 Clinical positioning and indication of pembrolizumab

In representative clinical practice guidelines and representative textbooks for clinical oncology in and outside

Japan, pembrolizumab therapy for NSCLC is currently described as follows:

[Clinical practice guidelines]

• NCCN guidelines (v. 6. 2018):

Pembrolizumab in combination with a platinum + PEM is strongly recommended as the first-line

therapy for chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC.

Pembrolizumab in combination with CBDCA + PTX or CBDCA + nab-PTX is recommended as the

first-line therapy for chemotherapy-naïve patients with unresectable, advanced or recurrent SQ-

NSCLC.

Pembrolizumab monotherapy is strongly recommended as the first-line therapy for chemotherapy-

naïve patients with unresectable, advanced or recurrent PD-L1 positive (TPS ≥50%) NSCLC.

56


• US National Cancer Institute Physician Data Query (NCI PDQ) (version dated August 6, 2018):

Pembrolizumab in combination with a platinum + PEM is recommended as the first-line therapy for

chemotherapy-naïve patients with unresectable, advanced or recurrent NSQ-NSCLC.

Pembrolizumab monotherapy is recommended as the first-line therapy for chemotherapy-naïve

patients with unresectable, advanced or recurrent PD-L1 (TPS ≥50%) NSCLC.

• Japanese clinical practice guidelines (NSCLC):

Pembrolizumab monotherapy is strongly recommended as the first-line therapy for chemotherapy-

naïve patients with unresectable, advanced or recurrent PD-L1 (TPS ≥50%) NSCLC.

PMDA asked the applicant to explain about the clinical positioning and intended population of pembrolizumab

in the treatment of unresectable, advanced or recurrent NSCLC.


Based on the results of Studies 189, 407, and 042, the combination of pembrolizumab and chemotherapy can

be positioned as a therapeutic option for chemotherapy-naïve patients with advanced or recurrent NSCLC, and

pembrolizumab monotherapy can be positioned as a therapeutic option for chemotherapy-naïve patients with

advanced or recurrent PD-L1 positive (TPS ≥1%) NSCLC. Furthermore, pembrolizumab in combination with

chemotherapy was demonstrated to be effective and safe, regardless of PD-L1 expression status, in Studies 189

and 407; therefore, PD-L1 testing has little clinical significance in the combination therapy for NSCLC in

clinical practice, while PD-L1 testing is necessary when pembrolizumab is administered alone.

In view of the above, the indication of pembrolizumab for NSCLC can be changed to “unresectable, advanced

or recurrent NSCLC,” provided that the following precautionary statements are included in the “Precautions

for Indications” section of the package insert.



• The “Clinical Studies” section should be carefully read to understand the percentage of PD-L1-expressing

tumor cells (TPS). PD-L1 testing should be performed by a highly experienced pathologist or at a

laboratory facility using the approved in vitro diagnostic. When used as monotherapy, pembrolizumab

should be administered to patients whose tumors have been demonstrated to express PD-L1.




approval).

PMDA asked the applicant to explain the choice between pembrolizumab monotherapy and pembrolizumab in

combination with platinum-based chemotherapy in the treatment of chemotherapy-naïve patients with

advanced or recurrent NSCLC.

57



Based on the results of Studies 189 and 407, pembrolizumab should be used in combination with other

antineoplastic drugs, regardless of PD-L1 expression status, in the treatment of chemotherapy-naïve patients

with advanced or recurrent NSCLC. In addition, the possibility of pembrolizumab monotherapy should also be

considered in PD-L1 positive (TPS ≥1%) patients, in view of the results of Studies 024 and 042.

Although there have been no clinical data that compare the efficacy and safety of pembrolizumab monotherapy

with those of combination therapy with pembrolizumab and platinum-based chemotherapy, the combination

therapy, rather than the monotherapy, should be chosen in the treatment of PD-L1 positive (TPS ≥1%) patients,

except the monotherapy should be considered for patients who are intolerant to combination therapy with

pembrolizumab and platinum-based chemotherapy, and those who are unwilling to receive chemotherapy.

PMDA's view:

Considering that pembrolizumab is intended to be used by physicians with sufficient knowledge and experience

in cancer chemotherapy, the indication of pembrolizumab for NSCLC can be changed as the applicant has

proposed, provided that the PD-L1 expression status or other characteristics of the patients enrolled in Study

189, 407, or 042 are detailed in the “Clinical Studies” section of the package insert, and the following

statements are included in the “Precautions for Indications” section.



• When used as monotherapy, pembrolizumab should be administered to patients whose tumors have been

demonstrated to express PD-L1. The “Clinical Studies” section should be carefully read to understand the

percentage of PD-L1-expressing tumor cells (TPS). PD-L1 testing should be performed by a highly

experienced pathologist or at a laboratory facility using the approved in vitro diagnostic.




approval).

Since there have been no clinical data that compare the efficacy and safety of pembrolizumab monotherapy

with those of combination therapy with pembrolizumab and platinum-based chemotherapy in the treatment of

chemotherapy-naïve patients with unresectable, advanced or recurrent PD-L1 positive (TPS ≥1%) NSCLC,

differences in the use of pembrolizumab alone, versus that in combination with platinum-based chemotherapy

are currently unclear, and an appropriate treatment regimen should be chosen according to individual patient

conditions.

58


7.2.R.3.2 Efficacy and safety of pembrolizumab by PD-L1 expression status (TPS), and intended

patient population

Pembrolizumab is an antibody against human PD-1. PMDA asked the applicant to explain the efficacy and

safety of pembrolizumab according to the expression status of PD-L1, a ligand of PD-1, and to describe the

intended patient population of pembrolizumab.


PD-L1 expression status in tumor tissue samples (TPS) were measured using “Clinical Trial Assay ‘Dako’

(Dako Japan Co., Ltd.)” in Study 189 or using “PD-L1 IHC 22C3 pharmDx assay ‘Dako’ (Dako Japan Co.,

Ltd.)” in Studies 407 and 042, to evaluate the efficacy and safety of pembrolizumab by TPS (cutoff values: 1%

or 50% in Studies 189 and 407, and 20% or 50% in Study 042) in (a) Study 189, (b) Study 407, and (c) Study

042, as described below. The results indicated that pembrolizumab in combination with platinum-based

chemotherapy can be recommended, regardless of PD-L1 expression status, for the treatment of chemotherapy-

naïve patients with advanced or recurrent NSCLC, and that pembrolizumab monotherapy can be recommended,

regardless of PD-L1 expression status, for the treatment of chemotherapy-naïve patients with advanced or

recurrent PD-L1 positive (TPS ≥1%) NSCLC.

(a) Study 189:

The efficacy of pembrolizumab was evaluated by TPS (cutoff value, 1% or 50%) in patients with NSQ-NSCLC

with evaluable TPS data. Table 27 and Figure 16 show the OS in these patients by TPS (cutoff value, 1% or

50%) (data cutoff date: November 8, 2017).

In both the PD-L1 positive and negative subgroups, based on each cutoff value, patients receiving

pembrolizumab showed a prolonged OS, compared with those receiving placebo. Pembrolizumab is thus

expected to be effective, regardless of PD-L1 expression status, in patients with NSQ-NSCLC.

Table 27. Efficacy of pembrolizumab by TPS in tumor tissue samples (Study 189)

PD-L1 expression

Treatment N OS




TPS <1% Pembrolizumab 127 15.2 [12.3, —]

0.59 [0.38, 0.92] 0.3274

Placebo 63 12.0 [7.0, —]

TPS ≥1% Pembrolizumab 260 — [—,—]

0.47 [0.34, 0.66] Placebo 128 11.3 [8.7, —]

TPS <50% Pembrolizumab 255 — [15.2, —]

0.57 [0.41, 0.79] 0.2294

Placebo 121 12.1 [8.7, -]

TPS ≥50% Pembrolizumab 132 — [—,—]

0.42 [0.26, 0.68] Placebo 70 10.0 [7.5, —]

—, Not reached; *, Cox regression stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a platinum (CDDP, CBDCA), and smoking history (present, absent)

59


Figure 16. Kaplan-Meier curves of OS by TPS (data cutoff date: November 8, 2017)

(Top left, TPS <1%; top right, TPS ≥1%; bottom left, TPS <50%; bottom right, TPS ≥50%)

In the pembrolizumab group of Study 189, the incidences of adverse events of any grade were 100%, both in

patients with TPS <1% and in those with TPS ≥1%. The incidences of Grade ≥3 adverse events were 58.7% in

patients with TPS <1% and 71.5% in those with TPS ≥1%. The incidences of serious adverse events were

43.7% in patients with TPS <1% and 53.1% in those with TPS ≥1%. In addition, the incidences of adverse

events of any grade were 100%, both in patients with TPS <50% and in those with TPS ≥50%. The incidences

of Grade ≥3 adverse events were 62.8% in patients with TPS <50% and 76.0% in those with TPS ≥50%. The

incidences of serious adverse events were 46.2% in patients with TPS <50% and 57.4% in those with TPS

≥50%. These results suggested no clear differences in the safety of pembrolizumab, according to PD-L1

expression status in patients with NSQ-NSCLC.

(b) Study 407:

The efficacy of pembrolizumab was evaluated by TPS (cutoff value, 1% or 50%) in patients with SQ-NSCLC


50%) (data cutoff date: April 3, 2018).

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

60


In both the PD-L1 positive and negative subgroups, based on each cutoff value, patients receiving

pembrolizumab showed a prolonged OS, compared with those receiving placebo. Pembrolizumab is thus

expected to be effective, regardless of PD-L1 expression status, in patients with SQ-NSCLC.


PD-L1 expression

Treatment N OS





0.61 [0.38, 0.98] 0.9132

Placebo 99 10.2 [8.6, 13.8]

TPS ≥1% Pembrolizumab 176 16.6 [13.2, —]

0.65 [0.45, 0.92] Placebo 177 11.6 [8.9, -]


0.59 [0.42, 0.82] 0.7466

Placebo 203 11.1 [9.1, 13.8]

TPS ≥50% Pembrolizumab 73 — [11.3, —]

0.64 [0.37, 1.10] Placebo 73 — [7.4, —]

—, Not reached; *, Cox regression stratified by PD-L1 expression status (TPS ≥1%, <1%), choice of a taxane-based chemotherapy (PTX, nab-PTX), and geographic region (East Asia, non-East Asia)

Figure 17. Kaplan-Meier curves of OS by TPS (data cutoff date: April 3, 2018)

(Top left, TPS <1%; top right, TPS ≥1%; bottom left, TPS <50%; bottom right, TPS ≥50%)

In the pembrolizumab group of Study 407, the incidences of adverse events of any grade were 97.9% in patients

with TPS <1% and 98.3% in those with TPS ≥1%. The incidences of Grade ≥3 adverse events were 72.6% in

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

Placebo

Pembrolizumab

61


patients with TPS <1% and 67.0% in those with TPS ≥1%. The incidences of serious adverse events were

38.9% in patients with TPS <1% and 40.9% in those with TPS ≥1%. In addition, the incidences of adverse

events of any grade were 97.5% in patients with TPS <50% and 100% in those with TPS ≥50%. The incidences

of Grade ≥3 adverse events were 66.7% in patients with TPS <50% and 75.3% in those with TPS ≥50%. The

incidences of serious adverse events were 39.4% in patients with TPS <50% and 42.5% in those with TPS

≥50%. These results suggested no clear differences in the safety of pembrolizumab, according to the degree of

PD-L1 expression, also in patients with SQ-NSCLC.

(c) Study 042:

The efficacy of pembrolizumab was evaluated by TPS (cutoff value, 20% or 50%) in patients with NSCLC


50%) (data cutoff date: February 26, 2018).

In the PD-L1 positive subgroups based on both cutoff values, patients receiving pembrolizumab showed a

prolonged OS, compared with those receiving chemotherapy. Pembrolizumab is thus expected to be effective,

regardless of the degree of PD-L1 expression, in patients with PD-L1 positive (TPS ≥1%) NSCLC.


PD-L1 expression Treatment N OS




TPS ≥1% and <20% Pembrolizumab 224 12.9 [9.7, 19.3]

0.89 [0.71, 1.12] 0.3318

Chemotherapy 232 11.3 [9.7, 12.9]

TPS ≥20% Pembrolizumab 413 17.7 [15.3, 22.1]

0.77 [0.64, 0.92] Chemotherapy 405 13.0 [11.6, 15.3]

TPS ≥1% and <50% Pembrolizumab 338 13.4 [10.7, 18.2]

0.92 [0.77, 1.11] 0.0410

Chemotherapy 337 12.1 [11.0, 14.0]

TPS ≥50% Pembrolizumab 299 20.0 [15.4, 24.9]

0.69 [0.56, 0.85] Chemotherapy 300 12.2 [10.4, 14.2]

*, Cox regression stratified by ECOG PS (0, 1), tissue type (SQ, NSQ), and geographic region (East Asia, non-East Asia)

62


Figure 18. Kaplan-Meier curves of OS by TPS (data cutoff date: February 26, 2018)

(Top left, TPS ≥1% and <20%; top right, TPS ≥20%; bottom left, TPS ≥1% and <50%; bottom right, TPS ≥50%)

In the pembrolizumab group of Study 042, the incidences of adverse events of any grade were 96.0% in patients

with TPS ≥1% and <20%, and 95.9% in those with TPS ≥20%. The incidences of Grade ≥3 adverse events

were 47.8% in patients with TPS ≥1% and <20%, and 51.2% in those with TPS ≥20%. The incidences of

serious adverse events were 39.7% in patients with TPS ≥1% and <20%, and 41.3% in those with TPS ≥20%.

In addition, the incidences of adverse events of any grade were 95.6% in patients with TPS ≥1% and <50%,

and 96.3% in those with TPS ≥50%. The incidences of Grade ≥3 adverse events were 50.0% in patients with

TPS ≥1% and <50%, and 50.0% in those with TPS ≥50%. The incidences of serious adverse events were 40.2%

in patients with TPS ≥1% and <50%, and 41.3% in those with TPS ≥50%. These results suggested no clear

differences in the safety of pembrolizumab, according to the degree of PD-L1 expression, also in patients with

PD-L1 positive (TPS ≥1%) NSCLC.

PMDA's view:

PMDA accepted the applicant's explanation in general. However, the applicant should continue to collect

information on other possible predictors of response to pembrolizumab besides PD-L1 expression, and

appropriately provide any new findings to healthcare professionals.

Chemotherapy

Pembrolizumab

Chemotherapy

Pembrolizumab

Chemotherapy

Pembrolizumab

Chemotherapy

Pembrolizumab

Chemotherapy

Pembrolizumab

Chemotherapy

Pembrolizumab

Chemotherapy

Pembrolizumab

Chemotherapy

Pembrolizumab

63



In the present partial change application, the proposed dosage regimen for the treatment of NSCLC was “The

usual adult dosage is 200 mg of pembrolizumab (genetical recombination) infused intravenously over 30

minutes every 3 weeks” (unchanged from the previous approval). The following precautionary statements were

included in the proposed “Precautions for Dosage and Administration” section:

• When pembrolizumab is used with other antineoplastic drugs, the “Clinical Studies” section should be

understood sufficiently before starting the combination therapy.

• When pembrolizumab is used with other antineoplastic drugs, pembrolizumab should be administered first.

The package inserts of the concomitant antineoplastic drugs should be read thoroughly.

• Criteria for interruption and discontinuation of pembrolizumab in case of adverse reactions (unchanged

from the previous approval)


PMDA concluded that the proposed dosage regimen is acceptable, provided that the following precautionary

statements are included in the “Precautions for Dosage and Administration” section.


understood sufficiently before selecting concomitant antineoplastic drugs.



7.2.R.4.1 Dosage and administration of pembrolizumab

The applicant's rationale for the dosage regimen of pembrolizumab for the treatment of NSCLC:

In Studies 189, 407, and 042, a fixed dose of 200 mg every 3 weeks, rather than a body weight-based dose,

was selected based on data including the results of a PK/PD analysis (see “Review Report of Keytruda Injection

20 mg, Keytruda Injection 100 mg dated August 30, 2016” and “Review Report of Keytruda Injection 20 mg,

Keytruda Injection 100 mg dated November 15, 2016”). These studies demonstrated the clinically significant

efficacy [see Section 7.2.R.1] and acceptable tolerability of pembrolizumab [see Section 7.2.R.2]. The dosage

regimen of pembrolizumab for the treatment of NSCLC was therefore proposed to be “The usual adult dosage

is 200 mg of pembrolizumab (genetical recombination) infused intravenously over 30 minutes every 3 weeks.”

However, no clinical data are available as to the combination therapy with pembrolizumab and antineoplastic

drugs, except a platinum + PEM, in patients with NSQ-NSCLC, or as to the combination therapy, except

CBDCA + PTX or CBDCA + nab-PTX, in patients with SQ-NSCLC. Therefore, the following precautionary

statements are included in the “Precautions for Dosage and Administration” section, provided that information

on the antineoplastic drugs used in combination with pembrolizumab in the clinical studies is provided in the

“Clinical Studies” section.


understood sufficiently before starting the combination therapy.

64


• When pembrolizumab is used with other antineoplastic drugs, pembrolizumab should be administered first.

The package inserts of all concomitant antineoplastic drugs should be read thoroughly.

PMDA's view:

The applicant's explanation is acceptable. However, the following proposed precautionary statements are

general matters requiring no special caution, which are not necessary to include in the “Precautions for Dosage

and Administration” section: “When pembrolizumab is used with other antineoplastic drugs, pembrolizumab

should be administered first,” and “The package inserts of all concomitant antineoplastic drugs should be read

thoroughly.”


Post-marketing investigations in patients with NSCLC will be described in Section 7.3.R.4, together with those

in patients with malignant melanoma and those with MSI-High solid tumors.

7.3 Data on the treatment of MSI-High solid tumors and outline of the review conducted by PMDA

The applicant submitted efficacy and safety evaluation data, in the form of result data from 2 global phase II

studies (Table 30).

Table 30. Clinical studies on efficacy and safety

Data type

Region Study



Dosage regimen Key

endpoints

Evaluation data

Global

Study 164

Ⅱ

Cohort A: Chemotherapy-treated patients with unresectable, advanced or recurrent dMMR or MSI-High (PCR) colorectal cancer

61 Intravenous pembrolizumab 200 mg every 3 weeks

Efficacy Safety

Study 158

Ⅱ

Chemotherapy-treated patients with unresectable, advanced or recurrent MSI-High solid tumors (except colorectal cancer)

94 Intravenous pembrolizumab 200 mg every 3 weeks

Efficacy Safety

A summary of these clinical studies is presented below. Major adverse events other than deaths reported in the

studies are detailed in Section “7.4.3 Adverse events reported in clinical studies in patients with MSI-High

solid tumors.”



7.3.1.1.1 Global phase II study (CTD 5.3.5.2.1, Study 164, ongoing since ** 20** [data cutoff date:

February 10, 2017])

An open-label, uncontrolled study was conducted at 21 sites in 9 countries, including Japan, to evaluate the

efficacy and safety of pembrolizumab in chemotherapy-treated12) patients with unresectable, advanced or

12) When the study was planned, patients who had been treated with ≥2 chemotherapy regimens, including a fluoropyrimidine,

oxaliplatin (L-OHP), or irinotecan hydrochloride hydrate (CPT-11) were to be eligible, assuming FOLFOX and FOLFIRI as commonly used standard therapies. However, the study protocol was later amended to enroll patients who had received

65


recurrent dMMR13) or MSI-High14) colorectal cancer. The study comprises Cohort A (target sample size, 60

patients) and Cohort B (target sample size, 60 patients).15) The present review report describes only data from

Cohort A.

Patients received an intravenous dose of pembrolizumab 200 mg every 3 weeks for a total of 35 cycles, or until

disease progression was observed or a withdrawal criterion was met.

All 61 enrolled patients (including 7 Japanese patients) were treated with pembrolizumab and were included

in both the efficacy analysis and safety analysis.

The primary endpoint was the response rate as assessed centrally using RECIST Ver. 1.1, with a threshold

response rate set at 15%.16) An interim analysis for efficacy was planned when the 40th patient enrolled had

been followed up for 18 weeks, and the final analysis was planned when the last patient enrolled had been

followed up for 6 months. A Lan-DeMets α spending function of the O’Brien-Fleming type was to be used to

control the type I error associated with the interim analysis. However, because of *********************,

an interim analysis was performed when all of the enrolled patients had been followed up for ≥18 weeks (data

cutoff date: June 3, 2016), with no multiplicity adjustment. The final analysis was performed (data cutoff date:

August 3, 2016). However, *****************************************************, suggesting

*****************************************, the originally planned final analysis was regarded as a

supportive analysis, and an additional analysis was newly planned (Protocol Amendment Version 4 [dated **

**, 20**]).

The results of the response rate as assessed centrally using RECIST Ver. 1.1 (the primary endpoint) are shown

in Table 31 (data cutoff date: February 10, 2017).

chemotherapy regimens including a fluoropyrimidine, L-OHP, and CPT-11, considering that FOLFOXIRI was used as a standard therapy containing a fluoropyrimidine, L-OHP, and CPT-11 (Protocol Amendment Version 2 [dated ** **, 20**]).

13) A tumor tissue sample was classified as dMMR if the expression of ≥1 mismatch repair protein (MLH1, MSH2, MSH6, or PMS2) was absent by IHC assay.

14) A tumor tissue sample was classified as MSI-High, if the allelic size was changed in ≥2 microsatellite markers. 15) After the originally planned patient enrollment was completed, the study protocol was amended to add Cohort B, which comprised

patients with radically unresectable, advanced, or recurrent, dMMR or MSI-High colorectal cancer, who had been treated with either (a) a fluoropyrimidine + L-OHP or (b) a fluoropyrimidine + CTP-11 (target sample size, 60 patients). The original cohort, from the time of the planning of the study, was termed Cohort A (Protocol Amendment Version 3 [dated ** **, 20**]).

16) The threshold response rate of 15% was based on the results of a clinical study of regorafenib in patients with radically unresectable, advanced or recurrent colorectal cancer that had progressed after chemotherapies containing a fluoropyrimidine, L-OHP, CPT-11, and bevacizumab (with cetuximab or panitumumab, if KRAS wild-type) (Lancet 2013; 381:303-312).

66


Table 31. Best overall response and response rate (RECIST Ver. 1.1, central assessment, efficacy analysis set [data cutoff date: February 10, 2017])


Overall N = 61

Japanese N = 7

CR 0 0 PR 17 (27.9) 2 (28.6) SD 14 (23.0) 2 (28.6) PD 28 (45.9) 3 (42.9) NE 2 (3.3) 0

Response (CR + PR) (Response rate [95% CI*] [%]) 17 (27.9 [17.1, 40.8]) 2 (28.6 [3.7, 71.0]) *, Exact test

The safety analysis indicated that 1 (a non-Japanese patient) of 61 patients (1.6%) died during the treatment

period or within 90 days after the last dose. The cause of death was aspiration, for which a causal relationship

to the study drug was denied.

7.3.1.1.2 Global phase II study (CTD 5.3.5.2.2, Study 158, ongoing since ** 20** [data cutoff date:

April 28, 2017])

An open-label, uncontrolled study was conducted at 43 sites in 15 countries, including Japan, to evaluate the

efficacy and safety of pembrolizumab in chemotherapy-treated17) patients with unresectable, advanced or

recurrent solid tumors.18) The study comprised Groups A to K, divided according to cancer type (target sample

size: approximately 200 to 1350 patients, in total).

Patients received an intravenous dose of pembrolizumab 200 mg every 3 weeks for a total of 35 cycles, or until

disease progression was observed or a withdrawal criterion was met.

Of 1151 patients enrolled in the study (across Groups A to K), 94 patients (including 7 Japanese patients) with

MSI-High solid tumors19) (7 in Group D, 2 in Group E, 1 in Group H, 1 in Group J, and 83 in Group K) were

treated with pembrolizumab and were included in both the efficacy analysis set and the safety analysis set.

The results for the response rate as assessed centrally using RECIST Ver. 1.1 (the primary endpoint) in Study

158 (Group K) are presented in Table 32 (data cutoff date: April 28, 2017).

17) When the study was planned, patients who were resistant or intolerant to standard therapies were to be eligible, irrespective of prior

chemotherapies. However, as the medical history required for enrollment was not clearly defined, the study protocol was amended to enroll patients who were resistant or intolerant to chemotherapies recommended as standard first-line therapies, as well as to surgeries and radiotherapies routinely used in clinical practice (Protocol Amendment Version 1 [dated ** **, 20**]).

18) The following patients were enrolled in each group: Group A, patients with anal cancer (squamous cell carcinoma); Group B, biliary carcinoma (adenocarcinoma of the gallbladder and bile duct), excluding tumors of the ampulla of Vater; Group C, patients with (well- or moderately-differentiated) neuroendocrine tumors derived from the lung, appendix, small intestine, colon, rectum, or pancreas; Group D, patients with endometrial cancer (excluding sarcoma and mesenchymal tumors); Group E, cervical cancer (squamous cell carcinoma); Group F, patients with vulvar carcinoma (squamous cell carcinoma); Group G; patients with small cell lung cancer; Group H, patients with mesothelioma; Group I, patients with thyroid cancer; Group J, patients with salivary gland cancer (excluding sarcoma and mesenchymal tumors); and Group K, advanced, MSI-High solid tumors (except colorectal cancer). Patients enrolled in Groups A to J had to have a PD-L1 level, a gene expression profile score by RNA analysis, and a tumor tissue sample which could be used for MSI biomarker analysis.

19) In patients enrolled in Groups A to J, MSI status (MSI-High) was assessed retrospectively.

67


Table 32. Best overall response and response rate (Group K) (RECIST Ver. 1.1, central assessment, efficacy analysis set [data cutoff date: April 28, 2017])


Overall N = 83

Japanese N = 3

CR 4 (4.8) 0 PR 25 (30.1) 0 SD 20 (24.1) 2 (66.7) PD 24 (28.9) 1 (33.3) NE 10 (12.0) 0

Response (CR + PR) (Response rate [95% CI*] [%]) 29 (34.9 [24.8, 46.2]) 0

*, Exact test

Table 33 shows the results for the response rate as assessed centrally using RECIST Ver 1.1, by cancer type in

patients with MSI-High solid tumors enrolled in the study (Groups A to K). The number of Japanese patients

in each group and their best overall response were as follows: (a) endometrial cancer [Group D], 1 with PR;

(b) cervical cancer [Group E], 1 with PD; (c) mesothelioma [Group H], 1 with PD; (d) salivary gland cancer

[Group J], 1 with PR; (e) gastric cancer [Group K], 1 with SD and 1 with PD; and, (f) small intestine carcinoma

[Group K], 1 with SD).

Table 33. Response rate by cancer type in patients with MSI-High solid tumors [Groups A to K)

(RECIST Ver. 1.1, central assessment, efficacy analysis set [data cutoff date: April 28, 2017])

Cancer type n (%) Responders (CR + PR)

(Response rate [%]) N = 94

Endometrial cancer 24 (25.5) 13 (54.2) Gastric cancer 13 (13.8) 6 (46.2)

Small intestine carcinoma 13 (13.8) 4 (30.8) Pancreatic carcinoma 10 (10.6) 1 (10.0)

Biliary carcinoma 9 (9.6) 2 (22.2) Adrenocortical carcinoma 3 (3.2) 1 (33.3)

Mesothelioma 3 (3.2) 0 Small cell lung cancer 3 (3.2) 2 (66.7)

Cervical cancer 2 (2.1) 1 (50.0) Neuroendocrine tumors 2 (2.1) 0

Thyroid cancer 2 (2.1) 0 Urothelial carcinoma 2 (2.1) 1 (50.0)

Brain tumor 1 (1.1) 0 Ovarian cancer 1 (1.1) 0 Prostate cancer 1 (1.1) 0

Retroperitoneal tumor 1 (1.1) 1 (100) Salivary gland cancer 1 (1.1) 1 (100)

Sarcoma 1 (1.1) 1 (100) Testicular tumor 1 (1.1) 0

Tonsil cancer 1 (1.1) 1 (100)

The safety analysis indicated that 6 (including no Japanese patients) of 94 patients (6.4%) died during the

treatment period or within 90 days after the last dose. The causes of death were death in 2 patients, and acute

respiratory failure, Cushing's syndrome, general physical health deterioration, and pneumonia in 1 patient each.

A causal relationship to the study drug was denied for all of the events.

68


7.3.R Outline of the review conducted by PMDA for the treatment of MSI-High solid tumor

7.3.R.1 Safety [For adverse events, see Section “7.4.3 Adverse events reported in clinical studies in

patients with MSI-High solid tumors.”]

PMDA's view:

As a result of its review described in the section below, PMDA considers that the adverse events that require

special attention after the administration of pembrolizumab are those identified as requiring attention at the

regulatory reviews for the approved indications7) and should also be closely monitored when pembrolizumab

is administered to chemotherapy-treated patients with MSI-High solid tumors.

Although attention should be paid to the above events, pembrolizumab is also tolerable when administered to

chemotherapy-treated patients with MSI-High solid tumors, as long as they are followed up by physicians with

sufficient knowledge and experience in cancer chemotherapy, through monitoring of adverse events,

differential diagnosis and patient management in anticipation of adverse reactions due to an excessive immune

response, drug interruption, or other appropriate actions.

7.3.R.1.1 Differences in safety profile between Japanese and non-Japanese patients

The applicant's explanation about the safety profile of pembrolizumab, based on the safety data from Studies

164 and 158:

Table 34 shows a summary of the safety data from Studies 164 and 158.

Table 34. Safety summary (Studies 164 and 158)

n (%)

Study 164 Study 158 N = 61 N = 94

All adverse events 60 (98.4) 91 (96.8) Grade ≥3 adverse events 36 (59.0) 41 (43.6) Adverse events leading to death 1 (1.6) 6 (6.4) Serious adverse events 29 (47.5) 32 (34.0) Adverse events leading to drug discontinuation 4 (6.6) 13 (13.8) Adverse events leading to drug interruption 17 (27.9) 27 (28.7)

In Study 164, adverse events of any grade reported with a ≥10% incidence were diarrhoea (34.4% [21 of 61

patients]), nausea (32.8% [20 of 61 patients]), abdominal pain (31.1% [19 of 61 patients]), fatigue (29.5% [18

of 61 patients]), vomiting (26.2% [16 of 61 patients]), pyrexia (24.6% [15 of 61 patients]), asthenia (23.0% [14

of 61 patients]), constipation, decreased appetite, and cough (19.7% [12 of 61 patients] each), anaemia and

arthralgia (18.0% [11 of 61 patients] each), oedema peripheral and pruritus (14.8% [9 of 61 patients] each),

and ALT increased, AST increased, back pain, headache, insomnia, and rash (11.5% [7 of 61 patients] each).

Grade ≥3 adverse events reported with a ≥2% incidence were abdominal pain, ALT increased, anaemia, and

ileus (6.6% [4 of 61 patients] each), AST increased, lipase increased, and pancreatitis (4.9% [3 of 61 patients]

each), and asthenia, blood ALP increased, dehydration, fatigue, hypertension, hyponatraemia, pulmonary

embolism, small intestinal obstruction, and urinary tract infection (3.3% [2 of 61 patients] each). Serious

adverse events reported with a ≥2% incidence were ileus (6.6% [4 of 61 patients]), and abdominal pain,

pulmonary embolism, pyrexia, small intestinal obstruction, and urinary tract infection (3.3% [2 of 61 patients]

69


each). Adverse events that led to drug interruption with a ≥2% incidence were ALT increased (6.6% [4 of 61

patients]), anaemia and AST increased (4.9% [3 of 61 patients] each), and blood ALP increased and pulmonary

embolism (3.3% [2 of 61 patients] each). No adverse events led to death or led to drug discontinuation with a

≥2% incidence.

In Study 158, adverse events of any grade reported with a ≥10% incidence were diarrhoea (28.7% [27 of 94

patients]), fatigue (27.7% [26 of 94 patients]), nausea (25.5% [24 of 94 patients]), asthenia (18.1% [17 of 94

patients]), constipation (17.0% [16 of 94 patients]), vomiting and pyrexia (16.0% [15 of 94 patients] each),

decreased appetite (14.9% [14 of 94 patients]), pruritus (13.8% [13 of 94 patients]), and anaemia (12.8% [12

of 94 patients]). Grade ≥3 adverse events reported with a ≥2% incidence were anaemia (6.4% [6 of 94 patients]),

GGT increased (5.3% [5 of 94 patients]), blood ALP increased (4.3% [4 of 94 patients]), sepsis (3.2% [3 of 94

patients]), and ascites, death, diarrhoea, fatigue, hepatic enzyme increased, hyponatraemia, muscular weakness,

pneumonia, urinary tract infection, and weight increased (2.1% [2 of 94 patients] each). Adverse event that led

to death with a ≥2% incidence was death (2.1% [2 of 94 patients]). Serious adverse events reported with a ≥2%

incidence were sepsis (3.2% [3 of 94 patients]), and death, pneumonia, and pneumonitis (2.1% [2 of 94 patients]

each). Adverse events that led to drug discontinuation with a ≥2% incidence were blood ALP increased and

death (2.1% [2 of 94 patients] each). Adverse events that led to drug interruption with a ≥2% incidence were

ALT increased, blood creatinine increased, GGT increased, hypothyroidism, pneumonitis, sepsis, and

thrombocytopenia (2.1% [2 of 94 patients] each).

The applicant's explanation about the differences in the safety profile of pembrolizumab between patients

enrolled in Study 164 and patients treated for the approved indications:

Adverse events of any grade reported with a ≥5% higher incidence in patients enrolled in Study 164 than in all

of the other patient populations were diarrhoea (MSI-High colorectal cancer, 34.4%; radically unresectable

malignant melanoma, 26.2%; PD-L1 positive NSCLC, 15.2%, cHL, 17.1%; radically unresectable urothelial

carcinoma, 16.2%), nausea (32.8%, 25.3%, 20.2%, 13.3%, 20.7%), abdominal pain (31.1%, 11.8%, 5.7%, 5.2%,

12.8%), vomiting (26.2%, 13.5%, 12.0%, 15.2%, 14.7%), asthenia (23.0%, 15.7%, 10.3%, 6.7%, 11.3%),

oedema peripheral (14.8%, 8.9%, 8.3%, 4.8%, 9.8%), AST increased (11.5%, 6.4%, 5.3%, 2.9%, 5.3%), and

ileus (8.2%, 0.4%, 0.1%, 0%, 0.4%). Grade ≥3 adverse events reported with a ≥2% higher incidence in patients

enrolled in Study 164 than in any of the other patient populations were abdominal pain (6.6%, 1.3%, 0.5%,

0.5%, 1.1%), ALT increased (6.6%, 1.1%, 0.8%, 0.5%, 1.1%), ileus (6.6%, 0.3%, 0.1%, 0%, 0.4%), AST

increased (4.9%, 0.8%, 0.8%, 0%, 2.3%), pancreatitis (4.9%, 0.2%, 0.2%, 0%, 0%), lipase increased (4.9%,

0.1%, 0%, 0.5%, 0.4%), asthenia (3.3%, 1.2%, 1.2%, 0%, 0.8%), blood ALP increased (3.3%, 0.7%, 0.8%,

0.5%, 1.1%), and small intestinal obstruction (3.3%, 0.1%, 0%, 0%, 0%). Serious adverse events reported with

a ≥2% higher incidence in patients enrolled in Study 164 than in any of the other patient populations were ileus

(6.6%, 0.4%, 0.1%, 0%, 0%), abdominal pain (3.3%, 1.0%, 0.4%, 0%, 0.8%), and small intestinal obstruction

(3.3%, 0.2%, 0%, 0%, 0%). Adverse events that led to drug interruption with a ≥2% higher incidence in patients

enrolled in Study 164 than in any of the other patient populations were ALT increased (6.6%, 1.0%, 1.1%,

0.5%, 0.8%), AST increased (4.9%, 1.0%, 1.2%, 1.0%, 0.8%), anaemia (4.9%, 0.8%, 0.8%, 0%, 0.4%),

pulmonary embolism (3.3%, 0.4%, 0.4%, 0%, 0.4%), and blood ALP increased (3.3%, 0.1%, 0.4%, 0%, 0%).

70


No adverse events led to death or led to drug discontinuation, with a ≥2% higher incidence in patients enrolled

in Study 164 than in any of the other patient populations.

The adverse event that had not been identified in patients treated for the approved indications but occurred in

≥2 patients in Study 164 was duodenal ulcer (2 patients). There were no adverse events leading to death, Grade

≥3 adverse events, serious adverse events, or adverse events leading to drug discontinuation or drug

interruption, which had not been identified in patients treated for the approved indication but reported in ≥2

patients in Study 164.

The above comparisons revealed no clear differences in the incidences of clinically notable adverse events

(e.g., serious adverse events) between patients enrolled in Study 164 and those treated for the approved

indications, although some adverse events occurred more frequently or were newly identified in patients

enrolled in Study 164. Thus, the safety of pembrolizumab does not clearly differ between patients enrolled in

Study 164 and those treated for the approved indications.

The applicant's explanation about the differences in the safety of pembrolizumab between Japanese and non-

Japanese patients, based on the safety data from Studies 164 and 158:

In Study 164, adverse events of any grade reported in ≥2 Japanese patients were anaemia in 3 patients, and

vomiting, weight increased, cough, and rash in 2 patients each. Grade ≥3 adverse events reported in Japanese

patients were anaemia, duodenal ulcer, gastrointestinal perforation, lower gastrointestinal haemorrhage,

pancreatitis, oedema peripheral, amylase increased, lipase increased, and tumour pain in 1 patient each. Serious

adverse events reported in Japanese patients were tumour pain and duodenal ulcer in 1 patient each. The adverse

event that led to drug discontinuation in Japanese patients was pneumonitis in 1 patient. Adverse events that

led to drug interruption in Japanese patients were anaemia, pancreatitis, amylase increased, and lipase increased

in 1 patient each. No Japanese patient died due to adverse events.

In Study 158, adverse events of any grade reported in ≥2 Japanese patients were constipation and pruritus in 3

patients each, and anaemia, diarrhoea, nausea, fatigue, pyrexia, blood ALP increased, GGT increased, and rash

in 2 patients each. Grade ≥3 adverse events reported in Japanese patients were anaemia and GGT increased in

2 patients each, and cataract, ascites, blood ALP increased, haemoglobin decreased, and fulminant type 1

diabetes mellitus in 1 patient each. Serious adverse events reported in Japanese patients were cataract, spinal

fracture, and fulminant type 1 diabetes mellitus in 1 patient each. Adverse event that led to drug discontinuation

in Japanese patients were blood ALP increased, GGT increased, and haemoglobin decreased in 1 patient each.

Adverse event that led to drug interruption in Japanese patients were cataract, malaise, pain, spinal fracture,

GGT increased, fulminant type 1 diabetes mellitus, hyperglycaemia, and tumour pain in 1 patient each. No

Japanese patient died due to adverse events.

Only limited clinical experience with pembrolizumab is available in Japanese patients with MSI-High solid

tumors, and there are limitations for an evaluation of the safety of pembrolizumab between Japanese and non-

Japanese patients, based only on the data from Studies 164 and 158. Nevertheless, no adverse events requiring

71


special attention in Japanese patients were identified, in view of the absence of serious adverse events reported

in ≥2 Japanese patients in both studies, as well as other findings.

PMDA's view:

Some adverse events occurred more frequently in patients enrolled in Study 164 than in patients treated for the

approved indications; however, most of these events were known adverse events of pembrolizumab, and the

adverse events that were newly identified in Study 164 were reported by only a limited number of patients.

Therefore, PMDA has concluded that pembrolizumab is also tolerable in patients with MSI-High solid tumors,

as long as they are followed up by physicians with sufficient knowledge and experience in cancer chemotherapy,

through monitoring of adverse events, differential diagnosis and patient management in anticipation of adverse

reactions due to an excessive immune response, drug interruption, or other appropriate actions.

PMDA has also concluded that the currently available data indicate no adverse events requiring special

attention in Japanese patients with MSI-High solid tumors, although clinical experience with pembrolizumab

in Japanese patients is limited.

7.3.R.2 Clinical positioning, efficacy, and indication

The proposed indication was “locally advanced or metastatic MSI-High cancers,” and the following statements

were included in the proposed “Precautions for Indications” section:

• Pembrolizumab should be administered to patients who have been demonstrated to have an advanced MSI-

High solid tumor by a highly experienced pathologist or at a laboratory facility using the approved in vitro

diagnostic.

• The efficacy and safety of pembrolizumab as a first-line therapy have not been established.

As a result of its review described in Sections “7.3.R.1 Safety” and the sections below, PMDA has concluded

that the appropriate indication of pembrolizumab should be “advanced or recurrent MSI-High solid tumors that

have progressed after cancer chemotherapy (limit the use to patients who are refractory or intolerant to standard

treatments),” and that the characteristics (e.g., cancer type) and response to treatment (e.g., best overall

response) of the patients enrolled in Study 158 should be detailed in the “Clinical Studies” section of the

package insert, and the following precautionary statements should be included in the “Precautions for

Indications” section.



diagnostic.

• In the treatment of colorectal cancer, the efficacy and safety of pembrolizumab have not been established

in patients with no prior therapy with a fluoropyrimidine + L-OHP + CTP-11.

• In the treatment of solid tumors (except colorectal cancer), the efficacy and safety of pembrolizumab as a

first-line therapy have not been established. In the second-line therapy for these solid tumors, standard

treatments, if available, should be chosen before pembrolizumab.

• The efficacy and safety of pembrolizumab in adjuvant therapy have not been established.

72



pembrolizumab, after carefully reading the “Clinical Studies” section to understand particularly for the

characteristics (e.g., cancer type) of patients enrolled in the clinical studies as well as a careful

consideration of the choice of alternative therapies.

7.3.R.2.1 Clinical positioning and efficacy of pembrolizumab in the treatment of MSI-High solid tumors

In major clinical practice guidelines and textbooks for clinical oncology in and outside Japan, pembrolizumab

therapy for MSI-High solid tumors is described as follows:

Representative clinical practice

guidelines and textbooks Statements

Colon cancer

NCCN guidelines (Colon cancer) (v.3.2018)

Pembrolizumab is recommended as the second- or third-line therapy for patients with radically unresectable, advanced or recurrent dMMR or MSI-High (PCR) colon cancer.

US NCI-PDQ (version dated April 6, 2018)

Results of Study 164, etc. demonstrated the efficacy of pembrolizumab in the treatment of radically unresectable, advanced or recurrent dMMR or MSI-High (PCR) colon cancer.

Japanese clinical practice guidelines (Large intestine carcinoma)

Anti-PD-L1 antibodies have been reported to be effective in the treatment of radically unresectable, advanced or recurrent dMMR or MSI-High (PCR) colorectal cancer.

Rectal cancer

NCCN guidelines (Rectal cancer) (v.3.2018)

Pembrolizumab is recommended as the second- or third-line therapy for patients with radically unresectable, advanced or recurrent dMMR or MSI-High (PCR) rectal cancer.

US NCI-PDQ (version dated June 7, 2018)

The results of Study 164, etc. demonstrated the efficacy of pembrolizumab in the treatment of dMMR or MSI-High (PCR) rectal cancer.

Japanese clinical practice guidelines (Large intestine carcinoma)

Anti-PD-L1 antibodies have been reported to be effective in the treatment of radically unresectable, advanced or recurrent dMMR or MSI-High (PCR) colorectal cancer.

Gastric cancer NCCN guidelines (Gastric cancer) (v.2.2018)

Pembrolizumab is a therapeutic option for the second-line therapy for patients with radically unresectable, advanced or recurrent dMMR or MSI-High (PCR) gastric cancer.

Esophageal and esophagogastric junction cancers

NCCN guidelines (Esophageal and esophagogastric junction cancers) (v.2.2018)

Pembrolizumab is a therapeutic option for the second-line or subsequent therapy for patients with recurrent or distant metastatic dMMR or MSI-High (PCR) esophageal and esophagogastric junction cancers.

Pancreatic carcinoma

NCCN guidelines (Pancreatic adenocarcinoma) (v.2.2018)

Pembrolizumab is recommended as the second-line therapy for patients with radically unresectable dMMR or MSI-High (PCR) pancreatic carcinoma.

Hepatobiliary cancer

NCCN guidelines (Hepatobiliary cancer) (v.3.2018)

Pembrolizumab is a therapeutic option for the second-line therapy for patients with dMMR or MSI-High (PCR) hepatobiliary cancer.

Uterine neoplasms

NCCN guidelines (Uterine neoplasms)(v.2.2018)

Pembrolizumab is a therapeutic option for patients with advanced or recurrent dMMR or MSI-High (PCR) endometrial cancer.

Cervical cancer NCCN guidelines (Cervical cancer) (v.1.2019)

Pembrolizumab is a therapeutic option for patients with advanced or recurrent dMMR or MSI-High (PCR) cervical cancer.

Ovarian cancer NCCN guidelines (Ovarian cancer) (v.2.2018)

Pembrolizumab is a therapeutic option for patients with recurrent dMMR or MSI-High (PCR) ovarian cancer.

Prostate cancer NCCN guidelines (Prostate cancer) (v.4.2018)

Pembrolizumab is recommended as the second-line therapy for patients with distant metastatic, dMMR or MSI-High (PCR) prostate cancer.

Bone cancer NCCN guidelines (Bone cancer) (v.1.2019)

Pembrolizumab is a therapeutic option for chemotherapy-treated patients with dMMR or MSI-High (PCR) osteosarcoma.

The applicant's explanation about the clinical positioning of pembrolizumab in the treatment of MSI-High solid

tumors, taking account of the molecular pathogenesis of MSI-High solid tumors:

In general, (a) dMMR and (b) MSI status is determined or classified as follows:

73


(a) The expression of 4 mismatch repair proteins (MLH1, MSH2, MSH6, and PMS2) in a tumor tissue sample

is observed by IHC assay. If any of the proteins are absent, the sample is classified as “dMMR.”

(b) DNA is extracted from a tumor tissue sample, and 5 microsatellites (repeated sequences of 1 to several

bases) from the DNA are amplified by PCR. If the number of base repeats is changed from that in normal

tissues (non-tumor tissues) in ≥2 of the 5 microsatellites, the sample is classified as “MSI-High (PCR).”

Microsatellites are repetitive sequences located in the genome, and are often replicated incorrectly due to a

failure of the DNA mismatch repair system during DNA replication (Cell Res 2008;18:85-98). The phenotype

of MSI-High is closely associated with a loss of function of mismatch repair proteins (MLH1, MSH2, MSH6,

and PMS2). In MSI-High (PCR) solid tumors as compared with solid tumors with no MSI, somatic cell

mutations are frequently detected, and a large burden of neoantigens, which can serve as the targets of cancer

antigen-specific T cells, are formed. The microenvironment of MSI-High solid tumors is characterized by an

abundance of activated cytotoxic T cells (Cell Rep 2016;15:857-65, etc.)

Furthermore, immunosuppressive signaling molecules (e.g., PD-1) are highly expressed by MSI-High solid

tumors, making them resistant to tumor rejection (Cancer Discov 2015;5:43-51, etc.). In view of such findings,

pembrolizumab, which is an immune checkpoint inhibitor, is expected to be effective in MSI-High solid tumors,

regardless of cancer type.

The applicant's explanation about the efficacy of pembrolizumab, also taking the above view into account:

In chemotherapy-treated patients with radically unresectable, advanced or recurrent solid tumors, the

achievement of response is expected to improve clinical symptoms associated with disease progression (J Clin

Oncol 2008;26:2311-9, etc.). As achieving a response is thus clinically significant in chemotherapy-treated

patients with unresectable, advanced or recurrent solid tumor, response rate was selected as the primary

endpoint in Studies 164 and 158.

In Study 164, the response rate as assessed centrally using RECIST Ver. 1.1 (the primary endpoint) was 27.9

(95% confidence interval [CI], 17.1-40.8) (data cutoff date, February 10, 2017) [see Section 7.3.1.1.1]. In Study

158 (Group K), the response rate as assessed centrally using RECIST Ver. 1.1 (the primary endpoint) was 34.9

(95% CI, 24.8-46.2) (data cutoff date, April 28, 2017) [see Section 7.3.1.1.2]. Thus, pembrolizumab was shown

to produce a clinically significant response rate in the target patient populations of Studies 164 and 158.

Pembrolizumab can be expected to be effective in chemotherapy-treated patients with advanced or recurrent

dMMR or MSI-High (PCR) solid tumors, who have no other standard treatment options, also in view of the

following findings.

• In a global phase III study to evaluate the efficacy and safety of regorafenib versus placebo in patients with

radically unresectable, advanced or recurrent colorectal cancer that had progressed after ≥2 chemotherapies

containing a fluoropyrimidine, L-OHP, CPT-11, and bevacizumab (with cetuximab or panitumumab, if

KRAS wild-type), the response rate as assessed by the investigator was 1.0% in the regorafenib group

(Lancet 2013;381:303-312).

74


• In a global phase III study to evaluate the efficacy and safety of the combination therapy with trifuridine +

tipiracil hydrochloride versus placebo in patients with radically unresectable, advanced or recurrent

colorectal cancer, who had received ≥2 prior chemotherapies and were unresponsive or intolerant to a

fluoropyrimidine, L-OHP, CPT-11, and bevacizumab (as well as to cetuximab or panitumumab, if KRAS

wild-type), the response rate as assessed by the investigator was 1.6% in the trifuridine + tipiracil

hydrochloride group (N Engl J Med 2015;372:1909-19).

At the start of Study 164, as *****************************************************************

******************************************** was anticipated by referring to *************** and

***************************************************************************************

****************************************, the final analysis was planned when the last enrolled patient

had been followed up for 6 months. However, ************************************* was observed at

the data cutoff on August 3, 2016 (the time point of the originally planned final analysis), suggesting

*****************************************************. Thus the analysis was to be conducted

************************************************************************, and the analysis as

of the data cutoff of February 10, 2017 was retrospectively determined to be used as the final analysis.

The results for the response rate obtained at the time point of the originally planned final analysis (data cutoff

date: August 3, 2016) are presented in Table 35. The lower bound of the 95% CI was lower than the pre-defined

threshold response rate (15.0%).

Table 35. Best overall response and response rate in Study 164

(RECIST Ver. 1.1, central assessment, efficacy analysis set [data cutoff date: August 3, 2016])


Overall N = 61

Japanese N = 7

CR 0 0 PR 15 (24.6) 2 (28.6) SD 16 (26.2) 2 (28.6) PD 28 (45.9) 3 (42.9) NE 2 (3.3) 0

Response (CR + PR) (Response rate [95% CI*] [%]) 15 (24.6 [14.5, 37.3]) 2 (28.6 [3.7, 71.0]) *, Exact test

PMDA's view:

The relationship between OS (the true endpoint) and the response rate has not been clarified in patients with

radically unresectable, advanced or recurrent solid tumors, and it is difficult to evaluate the survival benefit of

pembrolizumab in such patients, based on the results for the response rate (the primary endpoint) from Study

164 and Study 158 (Group K). In addition, the lower bound of the 95% CI for the response rate at the data

cutoff of August 3, 2016 (the time point of the originally planned final analysis) was below the threshold

response rate (15.0%) that had been prespecified. Therefore, the results of Studies 164 and 158 (Group K)

should be interpreted carefully.

75


Nevertheless, the applicant's explanation about the efficacy of pembrolizumab is understandable. PMDA has

concluded that the efficacy of pembrolizumab can be expected in chemotherapy-treated patients with advanced

or recurrent MSI-High solid tumors, who have no other standard treatment options, including Japanese patients,

based on the results for the response rate obtained at the time point of the originally planned final analysis (data

cutoff date: August 3, 2016) in Study 164, and in view of the following facts.

• No clear differences were found in the pharmacokinetics of pembrolizumab administered alone between

Japanese and non-Japanese patients (see “Review Report for Keytruda Injection 20 mg, Keytruda Injection

100 mg dated November 15, 2016”).

• There are no clear differences in the diagnosis and treatment of unresectable, advanced or recurrent

colorectal cancer between Japan and outside Japan.

7.3.R.2.2 Intended population and indication of pembrolizumab for the treatment of MSI-High solid

tumors, and MSI-High testing

The applicant's explanation about the intended population of pembrolizumab, proposed based on the results of

Studies 164 and 158 in the present partial change application:

The results of Study 164 [see Section 7.3.1.1.1] suggest that pembrolizumab can be positioned as a therapeutic

option for dMMR or MSI-High (PCR) colorectal cancer that has progressed after the conventional standard

chemotherapies. In addition, since Study 164 allowed the enrollment of patients who had received a first-line

therapy with FOLFOXIRI containing 5-FU, L-OHP, and CPT-11, pembrolizumab can be recommended as a

second-line or subsequent therapy option for patients with prior therapy with FOLFOXIRI. Currently, a global,

open-label, randomized phase III study (Study 177) is ongoing to evaluate the efficacy and safety of

pembrolizumab versus conventional standard chemotherapies in chemotherapy-naïve patients with

unresectable, advanced or recurrent, dMMR or MSI-High (PCR) colorectal cancer (target sample size, 300

patients).

Based on the results of Study 158 [see Section 7.3.1.1.2] and considering that pembrolizumab is an immune

checkpoint inhibitor and expected to be effective for MSI-High solid tumors, regardless of cancer type [see

Section 7.3.R.2.1], pembrolizumab can be positioned as a therapeutic option for chemotherapy-treated patients

with advanced or recurrent MSI-High solid tumors (except colorectal cancer) including cancer types that were

not investigated in Study 158, who have no other standard treatment options. In addition, pembrolizumab can

be positioned as a second-line therapy option for solid tumors (except colorectal cancer), that have progressed

after standard chemotherapies, in view of the following findings.

• In Study 158, the response rate of pembrolizumab was 37.8% in patients with 1 prior chemotherapy, 34.6%

in those with 2 prior chemotherapies, and 36.4% in those with 3 prior chemotherapies, indicating no

significant differences in the response rate of pembrolizumab according to the number of prior

chemotherapies.

76


• In Study 061,20) the response rate in MSI-High patients (15 in the pembrolizumab group and 12 in the PTX

group) was 46.7% in the pembrolizumab group and 16.7% in the PTX group, suggesting that

pembrolizumab was highly effective as compared with PTX, which is a standard second-line therapy

(Lancet 2018;392:123-133 suppl).

Based on the above, the proposed indication of pembrolizumab was “locally advanced or metastatic MSI-High

cancers,” and the following statement was included in the proposed “Precautions for Indications” section:

• The efficacy and safety of pembrolizumab as a first-line therapy have not been established.

Patients for pembrolizumab therapy should be selected using the “MSI test kit (FALCO),” and the use of the

kit should be stated in the “Precautions for Indications” section, in view of the following.

• In Study 164, which demonstrated the clinical benefit of pembrolizumab, patients who were categorized

as having dMMR by IHC assay, and those categorized as having MSI-High (PCR) by PCR were enrolled.

Later, the equivalence between these assay methods and the “MSI test kit (FALCO)” was assessed, using

samples from the patients enrolled in Study 164. The results showed a positive agreement rate of 100%,

and suggested that the patient group that was categorized as having dMMR by IHC assay or categorized

as MSI-High (PCR) by PCR was identical to the patient group that was classified as MSI-High by the “MSI

test kit (FALCO).” Therefore, patients eligible for pembrolizumab therapy can also be selected by the use

of the “MSI test kit (FALCO),” in the post-marketing setting.

PMDA's view:

PMDA accepted the applicant's explanation in general. However, the use of pembrolizumab before standard

therapies such as a fluoropyrimidine, L-OHP, and CTP-11 cannot be recommended in chemotherapy-naïve

patients with unresectable, advanced or recurrent dMMR or MSI-High (PCR) colorectal cancer, since no

confirmatory clinical data have been obtained from patients treated with pembrolizumab for MSI-High

colorectal cancer. Therefore, it should be clearly stated in the “Indications” section that pembrolizumab is

indicated for the treatment of patients who are not candidates for standard treatments. In addition, a

precautionary statement to the effect that the efficacy and safety of pembrolizumab have not been established

in patients with no prior therapy with a fluoropyrimidine, L-OHP, and CPT-11 should be included in the

“Precautions for Indications” section.

Similarly, for the treatment of MSI-High solid tumors (except colorectal cancer), it should be clearly stated in

the “Indications” section that pembrolizumab is indicated for patients who are not candidates for standard

treatments. In addition, precautionary statements to the effect that the efficacy and safety of pembrolizumab as

a first-line therapy have not been established, and that standard treatments, if available, should be chosen before

pembrolizumab for the second-line therapy should be included in the “Precautions for Indications” section,

since no confirmatory clinical data are available.

20) A global, open-label, randomized, phase III study in patients with unresectable, advanced, or recurrent gastric or esophagogastric

junction cancer that has progressed after a first-line chemotherapy containing a fluoropyrimidine or a platinum-based chemotherapy, to evaluate the efficacy of safety of pembrolizumab versus PTX

77


The efficacy of pembrolizumab was evaluated primarily based on the results for the response rate, with no

survival benefit data available, and the use of alternative treatments should be considered carefully. Therefore,

a precautionary statement to the effect that the appropriateness of pembrolizumab therapy should be carefully

determined after sufficiently considering the possible use of alternative treatments should be included in the

“Precautions for Indications” section.

Based on the above, PMDA has concluded that the appropriate indication of pembrolizumab should be

“advanced or recurrent MSI-High solid tumors that have progressed after cancer chemotherapy (limit the use

to patients who are refractory or intolerant to standard treatments)” and that the characteristics (e.g., cancer

type) and response to treatment (e.g., best overall response) of the patients enrolled in Study 158 should be

detailed in the “Clinical Studies” section of the package insert. In addition, the precautionary statements below

should be included in the “Precautions for Indications” section:



diagnostic.

• In the treatment of colorectal cancer, the efficacy and safety of pembrolizumab have not been established

in patients with no prior therapy with a fluoropyrimidine + L-OHP + CTP-11.

• In the treatment of solid tumors (except colorectal cancer), the efficacy and safety of pembrolizumab as a

first-line therapy have not been established. In the second-line therapy for these solid tumors, standard

treatments, if available, should be chosen before pembrolizumab.

• The efficacy and safety of pembrolizumab in adjuvant therapy have not been established.



characteristics (e.g., cancer type) of the patients enrolled in the clinical studies as well as a careful

consideration of the choice of alternative therapies.

7.3.R.2.3 Efficacy and safety of pembrolizumab by PD-L1 expression status (combined positive score

[CPS]), and the intended population

Pembrolizumab is an antibody against human PD-1. PMDA therefore asked the applicant to explain the

efficacy and safety of pembrolizumab according to the expression status (CPS) of PD-L1, a ligand of PD-1,

and to identify the intended population of pembrolizumab.

The applicant's explanation about the efficacy and safety of pembrolizumab according to the PD-L1expression

status (CPS):

MSI-High and PD-L1 are predictors of response to pembrolizumab with different characteristics (Cancer

Discov 2015;5:43-51).

78


In Study 158, PD-L1 expression status in tumor tissue samples were measured using “PD-L1 IHC 22C3

pharmDx assay ‘Dako’ (Dako Japan Co., Ltd.)” to evaluate the (a) efficacy and (b) safety of pembrolizumab,

by CPS.

(a) Efficacy:

The efficacy of pembrolizumab was assessed by CPS (cutoff value, 1), in patients with evaluable CPS data in

Study 158. Table 36 shows the response rates by CPS (cutoff value, 1).

As responders were present in both PD-L1-positive (CPS ≥1) and -negative (CPS <1) populations, the efficacy

of pembrolizumab can be expected, regardless of PD-L1 expression status.

Table 36. Efficacy by CPS (Study 158)

PD-L1 expression Responders/N Response rate [95% CI*] (%)

CPS <1 6/19 31.6 [12.6, 56.6] CPS ≥1 11/22 50.0 [28.2, 71.8]

*, Exact test

(b) Safety:

The incidences of adverse events in patients with CPS <1 were 94.7% (any grade), 52.6% (Grade ≥3), and

47.4% (serious adverse events), while the incidences of adverse events in patients with CPS ≥1 were 95.5%

(any grade), 45.5% (Grade ≥3), and 36.4% (serious adverse events).

The safety of pembrolizumab did not clearly differ between PD-L1 positive (CPS ≥1) and PD-L1 negative

(CPS <1) patients, indicating that pembrolizumab is tolerable, regardless of PD-L1 expression status.

The investigations in (a) and (b), above suggest that response to pembrolizumab is independent of PD-L1

expression status in patients with MSI-High solid tumors.

Based on the above, pembrolizumab therapy can be recommended, regardless of PD-L1 expression status, for

the treatment of chemotherapy-treated patients with advanced or recurrent MSI-High solid tumors who have

no other standard treatment options.

PMDA's view:

PMDA accepted the applicant's explanation in general. However, the applicant should continue to collect

information on possible predictors of response to pembrolizumab in addition to PD-L1 expression, and

appropriately provide any new findings to healthcare professionals.


In the present partial change application, the proposed dosage regimen for the treatment of MSI-High solid

tumors was “The usual adult dosage is 200 mg of pembrolizumab (genetical recombination) infused

intravenously over 30 minutes every 3 weeks” (unchanged from the previous approvals). In addition, the

79


following precautionary statements were included in the proposed “Precautions for Dosage and Administration”

section:

• The efficacy and safety of pembrolizumab in combination with other antineoplastic drugs have not been

established.



As a result of its review described in Sections “7.3.R.1 Safety,” “7.3.R.2 Clinical positioning, efficacy and

indication,” and the section below, PMDA concluded that the appropriate dosage regimen for the treatment of

MSI-High solid tumors is “The usual adult dosage is 200 mg of pembrolizumab (genetical recombination)

infused intravenously over 30 minutes every 3 weeks,” and that the proposed precautionary statements should

be included in the “Precautions for Dosage and Administration” section.

7.3.R.3.1 Dosage and administration of pembrolizumab

The applicant's rationale for the dosage regimen of pembrolizumab selected for the treatment of MSI-High

solid tumors:

In Studies 164 and 158, a fixed dose of 200 mg every 3 weeks, rather than a body weight-based dose, was

selected based on data including the results of a PK/PD analysis (see “Review Report for Keytruda Injection

20 mg, Keytruda Injection 100 mg dated August 30, 2016” and “Review Report for Keytruda Injection 20 mg,

Keytruda Injection 100 mg dated November 15, 2016”). As a result, Studies 164 and 158 demonstrated the

clinically significant efficacy [see Section 7.3.R.2] and acceptable tolerability [see Section 7.3.R.1] of

pembrolizumab. Both the proposed dosage and administration and the precautionary statements in the proposed

“Precautions for Dosage and Administration” section are based on the dosage regimen used in Studies 164 and

158.



The applicant's explanation on the post-marketing surveillance plans for (a) malignant melanoma, (b) NSCLC,

and (c) MSI-High solid tumors:

(a) The initiation of a post-marketing surveillance covering patients treated with pembrolizumab for

malignant melanoma immediately after the approval is not necessary, in view of the following facts.

• A certain amount of data have been collected as to the safety of pembrolizumab in patients with

malignant melanoma through the ongoing post-marketing surveillance, in patients treated with

pembrolizumab for the previously approved indication of “unresectable malignant melanoma.”

• The safety profile of pembrolizumab observed in Study 054 did not clearly differ from that in patients

treated for the approved indications [see Section 7.1.R.2.1].

• No particular safety concerns in Japanese patients have been raised in Study 054 [see Section 7.1.R.2.1].

• In view of the results of a PPK analysis and an investigation based on the data from Studies 001, 002,

and 006 [see Sections 6.2.1 and 7.1.R.4.1], the change in the dosage regimens of pembrolizumab for

80


the treatment of malignant melanoma from a body weight-based dose to a fixed dose is unlikely to

raise any particular safety concerns.

(b) The initiation of a post-marketing surveillance covering patients treated with pembrolizumab for NSCLC

immediately after the approval is not necessary, in view of the following facts.

• The post-marketing surveillance covering patients treated with pembrolizumab for the approved

indication of “unresectable, advanced or recurrent PD-L1 positive NSCLC” is ongoing, and a certain

amount of data have been collected from Japanese patients treated with pembrolizumab for NSCLC.

• No new adverse events requiring attention were identified from the results of Studies 189, 407, or 042

[see Section 7.2.R.2.1].

• No particular safety concerns in Japanese patients have been raised in Studies 189, 407, or 042 [see

Section 7.2.R.2.1].

(c) In Studies 164 and 158, the duration of response did not reach the median, and currently available long-

term efficacy data are limited. Furthermore, clinical experience in Japanese patients is also limited. In

view of these facts, the applicant plans to conduct post-marketing surveillance covering patients with MSI-

High solid tumors treated with pembrolizumab, with the target sample size and observation period

specified below. The purpose of the surveillance is to evaluate the long-term efficacy of pembrolizumab

in clinical practice.

• Based on available data including the results of Studies 164 and 158, an expected response rate of 35%

with a null hypothesis of 5% was assumed in patients treated with pembrolizumab in clinical practice.

To reject the null hypothesis at a level of significance of 0.025 (one-sided), 20 patients are needed.

Considering dropouts, the target sample size is 30 patients.

• The average time to response (median) in Studies 164 and 158 was 4 months. In addition, considering

the follow-up periods in Studies 164 and 158 (12 months), the observation period required to evaluate

the duration of response as an endpoint of the long-term efficacy of pembrolizumab (the objective of

the surveillance) is 16 months.

On the other hand, it is not necessary to initiate the post-marketing surveillance to evaluate the safety of

pembrolizumab in patients treated with pembrolizumab for MSI-High solid tumors in a clinical setting

immediately after the approval, in view of the following facts.

• The safety profile of pembrolizumab did not clearly differ between patients enrolled in Study 164 and

patients treated for the approved indications [see Section 7.3.R.1].

• Post-marketing surveillance in patients treated for the approved indications is ongoing, and a certain

amount of data have been collected from Japanese patients treated with pembrolizumab.

For (a) malignant melanoma, (b) NSCLC, and MSI-High colorectal cancer among (c) MSI-High solid tumors,

PMDA's view on the plan for colorectal cancer is as follows:

As a result of its review described in Sections “7.1.R.2 Safety,” “7.1.R.4 Dosage and administration,” and a

certain amount of safety information collected from Japanese patients in addition to its review described in

Section “7.2.R.2 Safety,” PMDA concluded that there is little need to conduct a new post-marketing

81


surveillance to collect safety information immediately after the approval, and safety information may be

collected through routine pharmacovigilance activities.

PMDA's view on the post-marketing surveillance plan for (c) MSI-High solid tumors (except colorectal cancer)

is as follows:

• Since the cancer types investigated in Study 158 were limited, post-marketing surveillance should be

conducted to evaluate the efficacy of pembrolizumab administered to Japanese patients with MSI-High

solid tumors (except colorectal cancer) in clinical practice.

• In clinical practice, pembrolizumab is also expected to be administered to patients with solid tumor types

other than those investigated in Study 158. Therefore, as much safety information as possible should be

collected, including adverse reactions that are listed as the important identified risks included under the

RMP, through a post-marketing surveillance in Japanese patients with MSI-High solid tumors (except

colorectal cancer).

• The proposed target sample size and observation period in the post-marketing surveillance in Japanese

patients who are treated with pembrolizumab for MSI-High solid tumors (except colorectal cancer) in

clinical practice should be reconsidered to ensure the appropriate evaluation of the efficacy of

pembrolizumab in patients with MSI-High solid tumors (except colorectal cancer).

Studies 164 and 158 are still ongoing. Efficacy information should be collected continuously from these studies,

and the obtained information should be promptly provided to healthcare professionals.

7.4 Adverse events reported in clinical studies

Among the clinical study data submitted for safety evaluation, data on death are presented in the sections on

“Evaluation data” [Sections 7.1.1, 7.2.1, and 7.3.1]. Other major adverse events are presented below.

7.4.1 Adverse events reported in a clinical study in patients with malignant melanoma

7.4.1.1 Global phase III study (Study 054)

Adverse events were reported in 475 of 509 patients (93.3%) in the pembrolizumab group and 453 of 502

patients (90.2%) in the placebo group. Adverse events for which a causal relationship to the study drug could

not be ruled out were reported in 396 of 509 patients (77.8%) in the pembrolizumab group and 332 of 502

patients (66.1%) in the placebo group. Table 37 shows the adverse events reported with a ≥15% incidence in

either treatment group.

82


Table 37. Adverse events with a ≥15% incidence

SOC PT (MedDRA ver. 20.1)

n (%) Pembrolizumab

N = 509 Placebo N = 502

All grades Grade ≥3 All grades Grade ≥3

All adverse events 475 (93.3) 158 (31.0) 453 (90.2) 96 (19.1) Gastrointestinal disorders

Diarrhoea 141 (27.7) 6 (1.2) 130 (25.9) 6 (1.2) Nausea 88 (17.3) 1 (0.2)

General disorders and administration site conditions

Fatigue 168 (33.0) 4 (0.8) 168 (33.5) 3 (0.6) Investigations

Weight increased 63 (12.4) 0 82 (16.3) 0 Musculoskeletal and connective tissue disorders

Arthralgia 79 (15.5) 6 (1.2) 72 (14.3) 0 Nervous system disorders

Headache 95 (18.7) 0 93 (18.5) 1 (0.2) Skin and subcutaneous tissue disorders

Pruritus 99 (19.4) 0 58 (11.6) 0 Vascular disorders

Hypertension 74 (14.5) 28 (5.5) 77 (15.3) 18 (3.6)

Serious adverse events were reported in 128 of 509 patients (25.1%) in the pembrolizumab group and 82 of

502 patients (16.3%) in the placebo group. Serious adverse events reported in ≥5 patients in the pembrolizumab

group were basal cell carcinoma in 17 patients (3.3%), colitis in 8 patients (1.6%), pneumonitis in 7 patients

(1.4%), squamous cell carcinoma in 6 patients (1.2%), and diarrhoea in 5 patients (1.0%). Serious adverse

events reported in ≥5 patients in the placebo group were basal cell carcinoma in 25 patients (5.0%), cellulitis

in 7 patients (1.4%), and malignant melanoma in situ in 6 patients (1.2%). A causal relationship to the study

drug could not be ruled out for the colitis in 8 patients, pneumonitis in 7 patients, and diarrhoea in 4 patients

in the pembrolizumab group, and cellulitis in 1 patient in the placebo group.

Adverse events led to drug discontinuation in 70 of 509 patients (13.8%) in the pembrolizumab group and 18

of 502 patients (3.6%) in the placebo group. Adverse events that led to drug discontinuation in ≥5 patients in

either treatment group were pneumonitis in 7 patients (1.4%), colitis in 6 patients (1.2%), and diarrhoea in 5

patients (1.0%) in the pembrolizumab group. A causal relationship to the study drug could not be ruled out for

all of these events.

7.4.2 Adverse events reported in clinical studies in patients with NSCLC

7.4.2.1 Japanese phase I study (Study 011)

7.4.2.1.1 Part B

Adverse events were reported in 6 of 6 patients (100%) in Cohort 1 and 6 of 6 patients (100%) in Cohort 2.

Adverse events for which a causal relationship to the study drug could not be ruled out were 6 of 6 patients

(100%) in Cohort 1 and 6 of 6 patients (100%) in Cohort 2. Adverse events reported with a ≥60% incidence in

either cohort were nausea and decreased appetite in 6 patients (100%) each, neutropenia in 5 patients (83.3%),

anaemia, leukopenia, constipation, malaise, and hiccups in 4 patients (66.7%) each in Cohort 1; and nausea in

5 patients (83.3%), and anaemia and leukopenia in 4 patients (66.7%) each in Cohort 2.

83


Serious adverse events were reported in 2 of 6 patients (33.3%) in Cohort 1 and 4 of 6 patients (66.7%) in

Cohort 2. The serious adverse event reported in ≥2 patients in either cohort was pulmonary embolism in 2

patients (33.3%) in Cohort 2. A causal relationship to the study drug could not be ruled out for the pulmonary

embolism in 1 patient.

Adverse events led to drug discontinuation in 1 of 6 patients (16.7%) in Cohort 1, and 4 of 6 patients (66.7%)

in Cohort 2. The adverse event that led to drug discontinuation in ≥2 patients in either cohort was pulmonary

embolism in 2 patients (33.3%) in Cohort 2. A causal relationship to the study drug could not be ruled out for

the pulmonary embolism in 1 patient.

7.4.2.1.2 Part C

Adverse events were reported in 8 of 8 patients (100%) in Cohort 1 and 6 of 6 patients (100%) in Cohort 2.

Adverse events for which a causal relationship to the study drug could not be ruled out were 8 of 8 patients

(100%) in Cohort 1 and 6 of 6 patients (100%) in Cohort 2. Adverse events reported with a ≥60% incidence in

either cohort were neutropenia, peripheral sensory neuropathy, and alopecia in 5 patients (62.5%) each in

Cohort 1, anaemia, leukopenia, neutropenia, and decreased appetite in 6 patients (100%) each, nausea and

alopecia in 5 patients (83.3%) each, and thrombocytopenia, diarrhoea, and peripheral sensory neuropathy in 4

patients (66.7%) each in Cohort 2.

Serious adverse events were reported in 3 of 8 patients (37.5%) in Cohort 1, and 3 of 6 patients (50.0%) in

Cohort 2. The serious adverse event reported in ≥2 patients in either cohort was lung infection in 2 patients

(33.3%) in Cohort 2. A causal relationship to the study drug could not be ruled out for the lung infection in 1

patient.

Adverse events led to drug discontinuation in 3 of 8 patients (37.5%) in Cohort 1 and 3 of 6 patients (50.0%)

in Cohort 2. The adverse event that led to drug discontinuation in ≥2 patients in either cohort was lung infection

in 2 patients (33.3%) in Cohort 2. A causal relationship to the study drug could not be ruled out for the lung

infection in 1 patient.




not be ruled out were reported in 372 of 405 patients (91.9%) in the pembrolizumab group and 183 of 202

patients (90.6%) in the placebo group. Table 38 shows adverse events reported with a ≥20% incidence in either

treatment group.

84




n (%) Pembrolizumab

N = 405 Placebo N = 202


All adverse events 404 (99.8) 272 (67.2) 200 (99.0) 133 (65.8) Blood and lymphatic system disorders

Anaemia 187 (46.2) 66 (16.3) 94 (46.5) 31 (15.3) Neutropenia 110 (27.2) 64 (15.8) 49 (24.3) 24 (11.9)

Gastrointestinal disorders Constipation 141 (34.8) 4 (1.0) 64 (31.7) 1 (0.5) Diarrhoea 125 (30.9) 21 (5.2) 43 (21.3) 6 (3.0) Nausea 225 (55.6) 14 (3.5) 105 (52.0) 7 (3.5) Vomiting 98 (24.2) 15 (3.7) 47 (23.3) 6 (3.0)


Asthenia 83 (20.5) 25 (6.2) 49 (24.3) 7 (3.5) Fatigue 165 (40.7) 23 (5.7) 77 (38.1) 5 (2.5)

Metabolism and nutrition disorders Decreased appetite 114 (28.1) 6 (1.5) 61 (30.2) 1 (0.5)

Respiratory, thoracic, and mediastinal disorders

Cough 87 (21.5) 0 57 (28.2) 0 Dyspnoea 86 (21.2) 15 (3.7) 52 (25.7) 11 (5.4)

Skin and subcutaneous tissue disorders Rash 82 (20.2) 7 (1.7) 23 (11.4) 3 (1.5)


202 patients (47.0%) in the placebo group. Serious adverse events reported in ≥10 patients in the

pembrolizumab group were febrile neutropenia and pneumonia in 23 patients (5.7%) each, diarrhoea in 15

patients (3.7%), thrombocytopenia in 13 patients (3.2%), anaemia and pneumonitis in 12 patients (3.0%) each,

and neutropenia in 10 patients (2.5%). Serious adverse events reported in ≥10 patients in the placebo group

were pneumonia in 17 patients (8.4%) and anaemia in 10 patients (5.0%). A causal relationship to the study

drug could not be ruled out for the febrile neutropenia in 21 patients, thrombocytopenia in 13 patients, diarrhoea

in 12 patient, pneumonitis in 11 patients, anaemia in 9 patients, neutropenia in 7 patient, and pneumonia in 4

patients in the pembrolizumab group, and anaemia in 10 patients and pneumonia in 1 patient in the placebo

group.


of 202 patients (14.9%) in the placebo group. Adverse events that led to drug discontinuation in ≥10 patients

in either treatment group were acute kidney injury and pneumonitis in 12 patients (3.0%) in the pembrolizumab

group. A causal relationship to the study drug could not be ruled out for the pneumonitis in 11 patients and

acute kidney injury in 10 patients.




not be ruled out were in 265 of 278 patients (95.3%) in the pembrolizumab group and 249 of 280 patients

85


(88.9%) in the placebo group. Table 39 shows adverse events reported with a ≥20% incidence in either

treatment group.



n (%) Pembrolizumab

N = 278 Placebo N = 280



Anaemia 148 (53.2) 43 (15.5) 145 (51.8) 57 (20.4) Neutropenia 105 (37.8) 63 (22.7) 92 (32.9) 69 (24.6) Thrombocytopenia 85 (30.6) 19 (6.8) 65 (23.2) 18 (6.4)

Gastrointestinal disorders Constipation 64 (23.0) 2 (0.7) 61 (21.8) 3 (1.1) Diarrhoea 83 (29.9) 11 (4.0) 65 (23.2) 6 (2.1) Nausea 99 (35.6) 3 (1.1) 90 (32.1) 4 (1.4)


Asthenia 60 (21.6) 6 (2.2) 59 (21.1) 10 (3.6) Fatigue 63 (22.7) 9 (3.2) 72 (25.7) 11 (3.9)

Metabolism and nutrition disorders Decreased appetite 68 (24.5) 6 (2.2) 82 (29.3) 5 (1.8)

Musculoskeletal and connective tissue disorders

Arthralgia 57 (20.5) 4 (1.4) 40 (14.3) 2 (0.7) Nervous system disorders

Peripheral neuropathy 57 (20.5) 3 (1.1) 45 (16.1) 2 (0.7) Skin and subcutaneous tissue disorders

Alopecia 128 (46.0) 1 (0.4) 102 (36.4) 3 (1.1)


280 patients (38.2%) in the placebo group. Serious adverse events reported in ≥10 patients in the

pembrolizumab group were pneumonia in 16 patients (5.8%) and febrile neutropenia in 15 patients (5.4%).

Serious adverse events reported in ≥10 patients in the placebo group were pneumonia in 17 patients (6.1%)

and febrile neutropenia in 10 patients (3.6%). A causal relationship to the study drug could not be ruled out for

the febrile neutropenia in 14 patients and pneumonia in 7 patients in the pembrolizumab group, and febrile

neutropenia in 9 patients and pneumonia in 3 patients in the placebo group.


of 280 patients (11.8%) in the placebo group. Adverse events that led to drug discontinuation in ≥5 patients in

either treatment group were autoimmune hepatitis, neutropenia, and pneumonitis in 5 patients (1.8%) each in

the pembrolizumab group. A causal relationship to the study drug could not be ruled out for the autoimmune

hepatitis and neutropenia in 5 patients each, and pneumonitis in 4 patients.



patients (98.5%) in the chemotherapy group. Adverse events for which a causal relationship to the study drug

could not be ruled out were reported in 399 of 636 patients (62.7%) in the pembrolizumab group and 553 of

86


615 patients (89.9%) in the chemotherapy group. Table 40 shows the adverse events reported with a ≥15%

incidence in either treatment group.



n (%) Pembrolizumab

N = 636 Chemotherapy

N = 615 All grades Grade ≥3 All grades Grade ≥3


Anaemia 99 (15.6) 17 (2.7) 259 (42.1) 92 (15.0) Gastrointestinal disorders

Constipation 77 (12.1) 0 130 (21.1) 1 (0.2) Nausea 74 (11.6) 3 (0.5) 196 (31.9) 7 (1.1) Vomiting 51 (8.0) 3 (0.5) 107 (17.4) 3 (0.5)


Fatigue 101 (15.9) 12 (1.9) 126 (20.5) 9 (1.5) Metabolism and nutrition disorders

Decreased appetite 110 (17.3) 11 (1.7) 131 (21.3) 9 (1.5) Respiratory, thoracic, and mediastinal disorders

Cough 99 (15.6) 1 (0.2) 65 (10.6) 2 (0.3) Dyspnoea 105 (16.5) 13 (2.0) 70 (11.4) 5 (0.8)

Skin and subcutaneous tissue disorders Alopecia 3 (0.5) 0 138 (22.4) 7 (1.1)


615 patients (30.4%) in the chemotherapy group. Serious adverse events reported in ≥10 patients in the

pembrolizumab group were pneumonia in 47 patients (7.4%), pneumonitis in 25 patients (3.9%), pulmonary

embolism in 15 patients (2.4%), pleural effusion in 14 patients (2.2%), and death in 10 patients (1.6%). Serious

adverse events reported in ≥10 patients in the chemotherapy group were pneumonia in 32 patients (5.2%),

anaemia in 17 patients (2.8%), febrile neutropenia in 15 patients (2.4%), and pulmonary embolism in 11

patients (1.8%). A causal relationship to the study drug could not be ruled out for the pneumonitis in 25 patients,

pleural effusion in 6 patients, pulmonary embolism in 2 patients, and pneumonia and death in 1 patient each in

the pembrolizumab group, and anaemia in 15 patients, pneumonia and febrile neutropenia in 13 patients each,

and pulmonary embolism in 4 patients in the chemotherapy group.


of 615 patients (14.5%) in the chemotherapy group. Adverse events that led to drug discontinuation in ≥10

patients in either treatment group were pneumonitis in 19 patients (3.0%) and death in 10 patients (1.6%) in

the pembrolizumab group. A causal relationship to the study drug could not be ruled out for the pneumonitis

in 19 patients and death in 1 patient.

7.4.2.5 Foreign phase I/II study (Study 021)

7.4.2.5.1 Cohort A

Adverse events were reported in 13 of 13 patients (100%) in the 2 mg/kg group and 12 of 12 patients (100%)

in the 10 mg/kg group. Adverse events for which a causal relationship to the study drug could not be ruled out

87


were reported in 13 of 13 patients (100%) in the 2 mg/kg group and 12 of 12 patients (100%) in the 10 mg/kg

group. Adverse events reported with a ≥50% incidence in the 2 mg/kg group were constipation and fatigue in

8 patients (61.5%) each, nausea, decreased appetite, cough, and alopecia in 7 patients (53.8%) each. Adverse

events reported with a ≥50% incidence in the 10 mg/kg group were fatigue in 8 patients (66.7%), constipation

and alopecia in 7 patients (58.3%) each.

Serious adverse events were reported in 7 of 13 patients (53.8%) in the 2 mg/kg group and 5 of 12 patients

(41.7%) in the 10 mg/kg group. Serious adverse events reported in ≥2 patients in the 2 mg/kg group were

pneumonia in 3 patients (23.1%) and febrile neutropenia in 2 patients (15.4%). Serious adverse event reported

in ≥2 patients in the 10 mg/kg group was pneumonia in 2 patients (16.7%). A causal relationship to the study

drug could not be ruled out for the febrile neutropenia in 2 patients in the 2 mg/kg group.

Adverse events led to drug discontinuation in 2 of 13 patients (15.4%) in the 2 mg/kg group and 1 of 12 patients

(8.3%) in the 10 mg/kg group. No adverse events led to drug discontinuation in ≥2 patients in either treatment

group.

7.4.2.5.2 Cohort C

Adverse events were reported in 12 of 12 patients (100%) in the 2 mg/kg group and 12 of 12 patients (100%)

in the 10 mg/kg group. Adverse events for which a causal relationship to the study drug could not be ruled out

were reported in 12 of 12 patients (100%) in the 2 mg/kg group and 12 of 12 patients (100%) in the 10 mg/kg

group. Adverse events reported with a ≥50% incidence in the 2 mg/kg group were constipation in 8 patients

(66.7%), nausea in 7 patients (58.3%), and fatigue in 6 patients (50.0%). Adverse events reported with a ≥50%

incidence in the 10 mg/kg group were fatigue in 11 patients (91.7%), constipation in 7 patients (58.3%), and

anaemia in 6 patients (50.0%).

Serious adverse events were reported in 6 of 12 patients (50.0%) in the 2 mg/kg group and 7 of 12 patients

(58.3%) in the 10 mg/kg group. The serious adverse event reported in ≥2 patients in the 2 mg/kg group was

anaemia in 2 patients (16.7%). Serious adverse events reported in ≥2 patients in the 10 mg/kg group were

chronic obstructive pulmonary disease and death in 2 patients (16.7%) each. A causal relationship to the study

drug could not be ruled out for the anaemia in 1 patient in the 2 mg/kg group.

Adverse events led to drug discontinuation in 4 of 12 patients (33.3%) in the 2 mg/kg group and 3 of 12 patients

(25.0%) in the 10 mg/kg group. The adverse event that led to drug discontinuation in ≥2 patients in either

treatment group was rash in 2 patients (16.7%) in the 10 mg/kg group. A causal relationship to the study drug

could not be ruled out for both of these events.

7.4.2.5.3 Cohort G

Adverse events were reported in 59 of 59 patients (100%) in the pembrolizumab + CBDCA + PEM group and

61 of 62 patients (98.4%) in the CBDCA + PEM group. Adverse events for which a causal relationship to the

study drug could not be ruled out were reported in 55 of 59 patients (93.2%) in the pembrolizumab + CBDCA

88


+ PEM group and 57 of 62 patients (91.9%) in the CBDCA + PEM group. Adverse events reported with a

≥50% incidence in the pembrolizumab + CBDCA + PEM group were nausea in 44 patients (74.6%), fatigue

in 43 patients (72.9%), and constipation in 32 patients (54.2%). Adverse events reported with a ≥50% incidence

in the CBDCA + PEM group were anaemia and nausea in 36 patients (58.1%) each, and fatigue in 34 patients

(54.8%).

Serious adverse events were reported in 29 of 59 patients (49.2%) in the pembrolizumab + CBDCA + PEM

group and 20 of 62 patients (32.3%) in the CBDCA + PEM group. Serious adverse event reported in ≥3 patients

in either treatment group were acute kidney injury in 6 patients (10.2%), cellulitis in 4 patients (6.8%), and

pleural effusion in 3 patients (5.1%) in the pembrolizumab + CBDCA + PEM group. A causal relationship to

the study drug could not be ruled out for the acute kidney injury in 3 patients and cellulitis in 1 patient.

Adverse events led to drug discontinuation in 10 of 59 patients (16.9%) in the pembrolizumab + CBDCA +

PEM group and 9 of 62 patients (14.5%) in the CBDCA + PEM group. Adverse events that led to drug

discontinuation in ≥2 patients in either treatment group were acute kidney injury in 3 patients (5.1%) in the

pembrolizumab + CBDCA + PEM group, and fatigue and stomatitis in 2 patients (3.2%) each in the CBDCA

+ PEM group. A causal relationship to the study drug could not be ruled out for all the events.

7.4.3 Adverse events reported in clinical studies in patients with MSI-High solid tumors

7.4.3.1 Global phase II study (Study 164)

Adverse events were reported in 60 of 61 patients (98.4%), and adverse events for which a causal relationship

to the study drug could not be ruled out were reported in 35 of 61 patients (57.4%). Table 41 shows adverse

events reported with a ≥15% incidence.

89




n (%)

N = 61

All grades Grade ≥3

All adverse events 60 (98.4) 36 (59.0) Blood and lymphatic system disorders

Anaemia 11 (18.0) 4 (6.6) Gastrointestinal disorders

Abdominal pain 19 (31.1) 4 (6.6) Constipation 12 (19.7) 0 Diarrhoea 21 (34.4) 0 Nausea 20 (32.8) 0 Vomiting 16 (26.2) 0

General disorders and administration site conditions Asthenia 14 (23.0) 2 (3.3) Fatigue 18 (29.5) 2 (3.3) Pyrexia 15 (24.6) 1 (1.6)

Metabolism and nutrition disorders Decreased appetite 12 (19.7) 0

Musculoskeletal and connective tissue disorders

Arthralgia 11 (18.0) 0 Respiratory, thoracic, and mediastinal disorders

Cough 12 (19.7) 0

Serious adverse events were reported in 29 of 61 patients (47.5%). Serious adverse events reported in ≥2

patients were ileus in 4 patients (6.6%), and abdominal pain, pulmonary embolism, pyrexia, small intestinal

obstruction, and urinary tract infection in 2 patients (3.3%) each. A causal relationship to the study drug was

denied for all the events.

Adverse events led to drug discontinuation in 4 of 61 patients (6.6%). Adverse events that led to drug

discontinuation were autoimmune arthritis, pneumonitis, decreased appetite, and aspiration in 1 patient (1.6%)

each. A causal relationship to the study drug could not be ruled out for the pneumonitis in 1 patient.

7.4.3.2 Global phase II study (Study 158)

Adverse events were reported in 91 of 94 patients (96.8%), and adverse events for which a causal relationship

to the study drug could not be ruled out were reported in 58 of 94 patients (61.7%). Table 42 shows adverse

events reported with a ≥15% incidence.

90




n (%)

N = 94

All grades Grade ≥3

All adverse events 91 (96.8) 41 (43.6) Gastrointestinal disorders

Constipation 16 (17.0) 1 (1.1) Diarrhoea 27 (28.7) 2 (2.1) Nausea 24 (25.5) 0 Vomiting 15 (16.0) 0

General disorders and administration site conditions Asthenia 17 (18.1) 1 (1.1) Fatigue 26 (27.7) 2 (2.1) Pyrexia 15 (16.0) 0

Serious adverse events were reported in 32 of 94 patients (34.0%). The serious adverse events reported in ≥2

patients were sepsis in 3 patients (3.2%), and death, pneumonia, and pneumonitis in 2 patients (2.1%) each. A

causal relationship to the study drug could not be ruled out for the pneumonitis in 2 patients.

Adverse events led to drug discontinuation in 13 of 94 patients (13.8%). The adverse events that led to drug

discontinuation in ≥2 patients were blood ALP increased and death in 2 patients (2.1%) each. A causal

relationship to the study drug could not be ruled out for the blood ALP increased in 2 patients.

8. Results of Compliance Assessment Concerning the New Drug Application Data and Conclusion

Reached by PMDA

8.1 PMDA's conclusion concerning the results of document-based GLP/GCP inspections and data

integrity assessment

The inspections and assessment are currently ongoing. The results of the inspections and assessment, and

PMDA's conclusion are reported in the Review Report (2).

8.2 PMDA's conclusion concerning the results of the on-site GCP inspection

The inspections and assessment are currently ongoing. The results of the inspections and assessment, and

PMDA's conclusion are reported in the Review Report (2).

9. Overall Evaluation during Preparation of the Review Report (1)

On the basis of the data submitted, PMDA has concluded that pembrolizumab has efficacy for the indications

described below, and that pembrolizumab has acceptable safety in view of its benefits, and therefore that

pembrolizumab therapy is clinically significant as a therapeutic option for each of these indications. PMDA

has also concluded that the dosage regimen of pembrolizumab for the treatment of unresectable malignant

melanoma can be changed from the body weight-based dose to a fixed dose.

• Pembrolizumab monotherapy for the adjuvant treatment of malignant melanoma


naïve patients with unresectable, advanced or recurrent NSCLC

91




• Pembrolizumab monotherapy for the treatment of chemotherapy-treated patients with advanced or

recurrent MSI-High solid tumors, who have no other standard treatment options

For the treatment of MSI-High solid tumors, PMDA considers that the indication and post-marketing

investigation items, including the conditions for approval, should be further discussed.

PMDA has concluded that pembrolizumab may be approved if pembrolizumab is considered to have no

particular problems based on comments from the Expert Discussion.

92


Review Report (2)

November 16, 2018

Product Submitted for Approval Brand Name Keytruda Injection 20 mg Keytruda Injection 100 mg Non-proprietary Name Pembrolizumab (genetical recombination) Applicant MSD K.K.

Date of Application March 30, 2018, May 10, 2018, August 9, 201821)


See Appendix

1. Content of the Review

Comments made during the Expert Discussion and the subsequent review conducted by the Pharmaceuticals

and Medical Devices Agency (PMDA) are summarized below. The expert advisors present during the Expert

Discussion were nominated based on their declarations etc. concerning the product submitted for marketing

approval, in accordance with the provisions of the Rules for Convening Expert Discussions etc. by

Pharmaceuticals and Medical Devices Agency (PMDA Administrative Rule No. 8/2008 dated December 25,

2008).

1.1 Safety

PMDA's conclusion:

As a result of its review described in Sections “7.1.R.2 Safety,” “7.2.R.2 Safety,” and “7.3.R.1 Safety” of the

Review Report (1), PMDA concluded that the adverse events that require special attention after the

administration of pembrolizumab are those identified as requiring attention at the regulatory reviews for the

approved indications22) and should also be closely monitored when pembrolizumab is administered for the

following indications.

(a) Pembrolizumab monotherapy for patients with resected malignant melanoma at high risk for recurrence23)

(b) Pembrolizumab in combination with platinum-based chemotherapy for the treatment of chemotherapy-

naïve patients with unresectable, advanced or recurrent non-small cell lung cancer (NSCLC)

21) Partial change applications for (a) a new indication for the treatment of MSI-High solid tumors and a modified dosage and

administration for the treatment of unresectable malignant melanoma, (b) a new indication, and a dosage and administration for the adjuvant treatment of malignant melanoma, and (c) a modified indication for the treatment of NSCLC were filed on (a) March 30, 2018, (b) May 10, 2018, and (c) August 9, 2018, respectively.

22) Gastrointestinal disorders, skin disorders, neurological disorders, hepatic function disorder, cholangitis sclerosing, eye disorders, endocrine disorders, renal impairment, interstitial lung disease (ILD), infusion-related reaction (IRR), pancreatitis, myositis, rhabdomyolysis, encephalitis and meningitis, myasthenia gravis, myocarditis, immune thrombocytopenic purpura, haemolytic anaemia, and pure red-cell aplasia (see “Review Report for Keytruda Injection 20 mg, Keytruda Injection 100 mg dated November 15, 2016”).

23) Patients with (a) stage IIIA (>1 mm lymph node metastasis), (b) stage IIIB, or (c) stage IIIC melanoma were defined as being at high risk for recurrence.

93


(c) Pembrolizumab monotherapy for the treatment of chemotherapy-naïve patients with unresectable,


(d) Pembrolizumab monotherapy for chemotherapy-treated patients with advanced or recurrent microsatellite

instability-high (MSI-High) solid tumors

Attention should be paid to the occurrence of the above-described adverse events; however, pembrolizumab is

also tolerable when administered for the above indications (a) to (d), as long as the patients are followed up by

physicians with sufficient knowledge and experience in cancer chemotherapy, through monitoring of adverse

events, differential diagnosis and patient management in anticipation of adverse reactions due to an excessive

immune response, drug interruption, or other appropriate actions.

At the Expert Discussion, the expert advisors supported PMDA's conclusion.

1.2 Efficacy, clinical positioning, and indications

PMDA's conclusion:

As a result of its review described in Sections “7.1.R.1 Efficacy,” “7.2.R.1 Efficacy,” and “7.3.R.2 Clinical

positioning, efficacy, and indication” of the Review Report (1), PMDA concluded the efficacy of

pembrolizumab in the treatment of (a) malignant melanoma, (b) NSCLC, and (c) MSI-High solid tumors as

follows:

(a) A global phase III study in patients with resected malignant melanoma at high risk for recurrence (Study

054) demonstrated the superiority of pembrolizumab to placebo in investigator-assessed RFS24) (the

primary endpoint); therefore, the efficacy of pembrolizumab has been demonstrated in the target patient

population of the study.

(b) Based on data, including the results of the following 3 global phase III studies (Studies 189, 407, and 042),

the efficacy of pembrolizumab has been demonstrated in the target patient populations of these studies.

• In a global phase III study in chemotherapy-naïve patients with unresectable, advanced or recurrent

non-squamous (NSQ)-NSCLC (Study 189) and a global phase III study in chemotherapy-naïve patients

with unresectable, advanced or recurrent squamous (SQ)-NSCLC (Study 407), pembrolizumab was

superior to placebo in OS (one of the primary endpoints).

• In a global phase III study in chemotherapy-naïve patients with unresectable, advanced or recurrent

PD-L1 positive (TPS ≥1%) NSCLC (Study 042), pembrolizumab was superior to chemotherapy in OS

(the primary endpoint).

(c) In a global phase II study in chemotherapy-treated patients with radically unresectable, advanced or

recurrent dMMR or MSI-High (PCR) colorectal cancer (Study 164) and a global phase II study in

chemotherapy-treated patients with radically unresectable, advanced or recurrent solid tumors (Study 158)

(Group K25)), the centrally assessed response rates (the primary endpoints) were 27.9 [95% CI, 17.1-40.8]

and 34.9 [95% CI, 24.8-46.2], respectively. Based on the study results, and taking into account the

24) RFS was defined as the time between the date of randomization and the date of first recurrence (local, regional, or distant metastasis)

or death, whichever occurs first. 25) Patients with MSI-High solid tumors (except colorectal cancer) were enrolled.

94


molecular pathogenesis of MSI-High solid tumors [see Section 7.3.R.2.1], a certain level of efficacy of

pembrolizumab has been demonstrated in the target study populations of the studies.

As a result of its review described in Sections “7.1.R.3 Clinical positioning and indication,” “7.2.R.3 Clinical

positioning and indication,” and “7.3.R.2 Clinical positioning, efficacy, and indication” of the Review Report

(1), PMDA concluded that the statements for (a) malignant melanoma, (b) NSCLC, and (c) MSI-High solid

tumors presented in the table below should be included in the “Indications” and “Precautions for Indications”

sections of the package insert.

Indications Precautions for Indications

(a) Malignant melanoma*1 • Appropriate patients should be selected based on a good understanding of the efficacy and safety of pembrolizumab, after carefully reading the “Clinical Studies” section.

(b) Unresectable, advanced or recurrent NSCLC*2

• The efficacy and safety of pembrolizumab in adjuvant therapy have not been established. • When used as a monotherapy, pembrolizumab should be administered to patients whose tumors

have been demonstrated to express PD-L1. The “Clinical Studies” section should be carefully read to understand the percentage of PD-L1-expressing tumor cells (TPS). PD-L1 testing should be performed by a highly experienced pathologist or at a laboratory facility using the approved in vitro diagnostic.

• Appropriate patients should be selected based on a good understanding of the efficacy and safety of pembrolizumab, after carefully reading the “Clinical Studies” section to understand the status of EGFR mutation or ALK fusion genes in patients enrolled in the clinical studies.

(c)

Advanced or recurrent microsatellite instability-high (MSI-High) solid tumors that have progressed after cancer chemotherapy (limit the use to patients who are refractory or intolerant to standard treatments)

• Pembrolizumab should be administered to patients who have been demonstrated to have an advanced MSI-High solid tumor by a highly experienced pathologist or at a laboratory facility using the approved in vitro diagnostic.

• In the treatment of colorectal cancer, the efficacy and safety of pembrolizumab have not been established in patients with no prior therapy with a fluoropyrimidine + L-OHP + CTP-11.

• In the treatment of solid tumors (except colorectal cancer), the efficacy and safety of pembrolizumab as a first-line therapy have not been established. In the second-line therapy for these solid tumors, standard treatments, if available, should be chosen before pembrolizumab.

• The efficacy and safety of pembrolizumab in adjuvant therapy have not been established. • Appropriate patients should be selected based on a good understanding of the efficacy and safety

of pembrolizumab, after carefully reading the “Clinical Studies” section to understand the characteristics (e.g., cancer type) of the patients enrolled in the clinical studies as well as carefully considering the choice of alternative therapies.

*1, Changed from the approved indication of “unresectable malignant melanoma” *2, Changed from the approved indication of “unresectable, advanced or recurrent PD-L1 positive NSCLC”

Discussion at the Expert Discussion:

Some expert advisors supported the above conclusions of the PMDA, while others made the following

comment:

• One suggestion is that a more detailed precautionary statement should be included in the “Precautions for

Indications” section when selecting patients for adjuvant treatment with pembrolizumab for malignant

melanoma.

PMDA's view:

Based on the comment of expert advisors, PMDA has concluded that the following precautionary statement

for the adjuvant treatment of malignant melanoma should be included in the “Precautions for Indications”

section.



disease stages of patients enrolled in the clinical studies.

95


In view of the above, PMDA instructed the applicant to include the above statement in the “Indications” and

“Precautions for Indications” sections of the package insert. The applicant agreed.

1.3 Dosage and administration

As a result of its review described in Sections “7.1.R.4 Dosage and administration,” “7.2.R.4 Dosage and

administration,” and “7.3.R.3 Dosage and administration” of the Review Report (1), PMDA concluded that the

following statements should be included in the “Dosage and administration” and “Precautions for Dosage and

Administration” sections of the package insert.

[Dosage and Administration]

• Malignant melanoma

The usual adult dosage is 200 mg of pembrolizumab (genetical recombination) infused intravenously

over 30 minutes every 3 weeks. For the adjuvant treatment of malignant melanoma, the maximum

duration of treatment is 12 months.

• Unresectable, advanced or recurrent non-small cell lung cancer, or advanced or recurrent MSI-High solid

tumors that have progressed after cancer chemotherapy (limit the use to patients who are refractory or

intolerant to standard treatments)

The usual adult dosage is 200 mg of pembrolizumab (genetical recombination) infused intravenously

over 30 minutes every 3 weeks.

[Precautions for Dosage and Administration]

• Malignant melanoma

The efficacy and safety of pembrolizumab in combination with other antineoplastic drugs have not

been established.

Criteria for interruption and discontinuation of pembrolizumab in case of adverse reactions (unchanged


• Unresectable, advanced or recurrent NSCLC

When pembrolizumab is used with other antineoplastic drugs, the “Clinical Studies” section should be


Criteria for interruption and discontinuation of pembrolizumab in case of adverse reactions (unchanged


• Advanced or recurrent MSI-High solid tumors that have progressed after cancer chemotherapy (limit the

use to patients who are refractory or intolerant to standard treatments)

The efficacy and safety of pembrolizumab in combination with other antineoplastic drugs have not

been established.

Criteria for the interruption and discontinuation of pembrolizumab in case of adverse reactions

(unchanged from the previous approval)


96


PMDA instructed the applicant to include the above statements in the “Dosage and Administration” and

“Precautions for Dosage and Administration” sections of the package insert. The applicant agreed.

1.4 Risk management plan (draft)

The applicant's explanation on the post-marketing surveillance plans for (a) malignant melanoma, (b) NSCLC,

and (c) MSI-High solid tumors:

For (a) and (b), there is little need to conduct post-marketing surveillance in association with the present partial

change application, immediately after the approval [see Section 7.3.R.4].

For (c), the applicant plans to conduct post-marketing surveillance covering patients treated with

pembrolizumab for MSI-High solid tumors, to evaluate the long-term efficacy of pembrolizumab in clinical

practice. The proposed target sample size is 30 patients, and the proposed observation period is 16 months. On

the other hand, it is not necessary to conduct post-marketing surveillance to evaluate the safety of

pembrolizumab in clinical practice, immediately after the approval [see Section 7.3.R.4].

As a result of its review described in Section “7.3.R.4 Post-marketing investigations” of the Review Report (1),

PMDA reached the following conclusions:

For (a) malignant melanoma, (b) NSCLC, and MSI-High colorectal cancer among (c) MSI-High solid tumors,

there is little need to conduct post-marketing surveillance in association with the present partial change

applications, immediately after the approval, and safety information may be collected through routine

pharmacovigilance activities.

For (c) MSI-High solid tumors (except colorectal cancer)

• Since only limited cancer types have been investigated, post-marketing surveillance should be conducted

to evaluate the efficacy of pembrolizumab administered to Japanese patients with MSI-High solid tumors

(except colorectal cancer) in clinical practice.The proposed target sample size and observation period

should be reconsidered to ensure the appropriate evaluation of the efficacy of pembrolizumab in Japanese

patients with MSI-High solid tumors (except colorectal cancer).

• As much safety information as possible should be collected from patients treated with pembrolizumab for

MSI-High solid tumors (except colorectal cancer), including the adverse reactions listed as important

identified risks under the risk management plan.

Studies 164 and 158 are still ongoing. Efficacy information should be collected continuously from these studies,

and the obtained information should be communicated promptly to healthcare professionals.


Based on the above review, PMDA instructed the applicant to reconsider the plan of the post-marketing

surveillance in patients treated with pembrolizumab for MSI-High solid tumors (except colorectal cancer).

97



• Based on the response rate of 37.2% (35 of 94 patients) observed in the pembrolizumab group of Study

158 and other relevant data, the number of patients required to achieve a statistical power of 95% is

calculated to be 20, when an expected response rate of 37.2% with a threshold response rate of 5% are

assumed, and a level of significance of 0.025 (one-sided) is used. Considering dropouts, the target sample

size is therefore 30 patients. Since the currently available data as to clinical experience with pembrolizumab

for MSI-High solid tumors (except colorectal cancer) are limited to several cancer types, ≥30 patients with

MSI-High solid tumors (except colorectal cancer) will be enrolled, and data will be collected from all

possible patients who are treated with pembrolizumab for MSI-High solid tumors (except colorectal

cancer) during the enrollment period. A total of 94 patients with MSI-High solid tumors are aimed to be

treated with pembrolizumab during the enrollment period.

• In Study 158, the maximum duration of pembrolizumab therapy to a response was 10.2 months. Based on

this, and considering the time needed to confirm best overall response, the observation period will be 12

months.

• The adverse reactions specified as important identified risks under the risk management plan (excluding

“Use in patients with a history of organ transplantation [including hematopoietic stem cell

transplantation]”) will be selected as the safety specifications for the surveillance, and information

regarding these adverse reactions will be collected.

PMDA accepted the applicant's response.

In view of the above discussions, PMDA has concluded that the risk management plan (draft) for

pembrolizumab should include the safety and efficacy specifications presented in Table 43, and that the

applicant should conduct additional pharmacovigilance activities, efficacy investigations or studies, and the

risk minimization activities presented in Tables 44 and 45.

98


Table 43. Safety and efficacy specifications in the risk management plan (draft) Safety specifications

Important identified risks Important potential risks Important missing information • ILD • Colitis and severe diarrhoea • Hepatic function disorder and cholangitis sclerosing • Renal impairment (tubulointerstitial nephritis, etc.) • Endocrine disorders (pituitary dysfunction, thyroid dysfunction,

and adrenal dysfunction) • Type 1 diabetes mellitus • Uveitis • Myositis and rhabdomyolysis • Pancreatitis • Neurological disorders (Guillain-Barre syndrome, etc.) • Severe skin disorders (oculomucocutaneous syndrome, erythema

multiforme, pemphigoid, etc.) • Encephalitis and meningitis • Myasthenia gravis • Myocarditis • Immune thrombocytopenic purpura • Haemolytic anaemia • Pure red-cell aplasia • Infusion related reaction (IRR) • Use in patients with a history of organ transplantation (including

hematopoietic stem cell transplantation)

• Increased risk of severe comorbidities associated with allogenic hematopoietic stem cell transplantation after pembrolizumab therapy (hematological malignancies)

• Embryonic/fetal toxicities

None

Efficacy specification • Efficacy in patients with unresectable malignant melanoma in clinical practice • Efficacy in patients with unresectable, advanced or recurrent PD-L1 positive NSCLC in clinical practice • Efficacy in patients with relapsed or refractory cHL in clinical practice • Efficacy in patients with advanced or recurrent MSI-High solid tumors (except colorectal cancer) that have progressed after cancer

chemotherapy in clinical practice Underline denotes specifications to be added after the present partial change approval.

Table 44. Summary of additional pharmacovigilance activities, efficacy investigations or studies, and additional

risk minimization activities included under the risk management plan (draft) Additional pharmacovigilance activities Efficacy investigations/studies Additional risk minimization activities

• Use-results survey in patients with unresectable malignant melanoma (all-case surveillance)

• Use-results survey in patients with unresectable, advanced or recurrent PD-L1 positive NSCLC (all-case surveillance)

• Use-results survey in patients with relapsed or refractory cHL (all-case surveillance)

• Use-results survey in patients with radically unresectable urothelial carcinoma that has progressed after cancer chemotherapy (all-case surveillance)

• Use-results survey in patients with advanced or recurrent MSI-High solid tumors (except colorectal cancer) that have progressed after cancer chemotherapy

• Post-marketing clinical studies (extension studies of Studies 041, 025, 010, 024, 087, 204, 045, 189, 407, and 042)

• Use-results survey in patients with unresectable malignant melanoma (all-case surveillance)

• Use-results survey in patients with unresectable, advanced or recurrent PD-L1 positive NSCLC (all-case surveillance)

• Use-results survey in patients with relapsed or refractory cHL (all-case surveillance)

• Use-results survey in patients with radically unresectable urothelial carcinoma that has progressed after cancer chemotherapy (all-case surveillance)

• Use-results survey in patients with advanced or recurrent MSI-High solid tumors (except colorectal cancer) that have progressed after cancer chemotherapy

• Post-marketing clinical studies (extension studies of Studies 041, 025, 010, 024, 087, 204, and 045)

• Organize and disseminate information for healthcare professionals

• Organize and disseminate information for patients

Underlines denote activities to be performed after the new indications are added.

99


Table 45. Outline of post-marketing surveillance plan (draft)

Objective To evaluate the efficacy of pembrolizumab in patients with advanced or recurrent MSI-High solid tumors (except colorectal cancer) that have progressed after cancer chemotherapy, and to collect safety information from those patients in clinical practice after the market launch

Survey method Central registration system

Population Patients with advanced or recurrent MSI-High solid tumors (except colorectal cancer) that have progressed after cancer chemotherapy

Observation period 12 months

Planned sample size

≥30 patients (Data will be collected from all possible patients who are treated with pembrolizumab for MSI-High solid tumors [except colorectal cancer] during the enrollment period [A total of 94 patients with MSI-High solid tumors are expected to be treated with pembrolizumab during the enrollment period]). )

Main survey items

Efficacy survey item: Efficacy of pembrolizumab in patients with advanced or recurrent MSI-High solid tumors (except colorectal cancer) that have progressed after cancer chemotherapy in clinical practice Safety survey items: ILD, colitis and severe diarrhoea, hepatic function disorder and cholangitis sclerosing, renal impairment (tubulointerstitial nephritis, etc.), endocrine disorders (pituitary dysfunction, thyroid dysfunction, and adrenal dysfunction), type 1 diabetes mellitus, uveitis, myositis and rhabdomyolysis, pancreatitis, neurological disorders (Guillain-Barre syndrome, etc.), severe skin disorders (oculomucocutaneous syndrome, erythema multiforme, pemphigoid, etc.), encephalitis and meningitis, myasthenia gravis, myocarditis, immune thrombocytopenic purpura, haemolytic anaemia, pure red-cell aplasia, and IRR Other main survey items: patient characteristics (e.g., age, sex, prior treatments, cancer type, disease stage classification), exposure to pembrolizumab, concomitant drugs and treatments, etc.

2. Results of Compliance Assessment Concerning the New Drug Application Data and Conclusion

Reached by PMDA

2.1 PMDA's conclusion on the results of document-based GLP/GCP inspections and data integrity

assessment

The new drug application data were subjected to a document-based compliance inspection and a data integrity

assessment in accordance with the provisions of the Act on Securing Quality, Efficacy and Safety of

Pharmaceuticals, Medical Devices, Regenerative and Cellular Therapy Products, Gene Therapy Products, and

Cosmetics. On the basis of the inspection and assessment, PMDA concluded that there were no obstacles to

conducting its review based on the application documents submitted.

2.2 PMDA's conclusion on the results of the on-site GCP inspection

The new drug application data (CTD 5.3.5.1.1 [malignant melanoma], CTD 5.3.5.1.1, CTD 5.3.5.1.5, CTD

5.3.5.1.6, and CTD 5.3.5.2.1 [NSCLC], and CTD 5.3.5.2.1 and CTD 5.3.5.2.2 [MSI-High solid tumors]) were

subjected to an on-site GCP inspection in accordance with the provisions of the Act on Securing Quality,

Efficacy and Safety of Pharmaceuticals, Medical Devices, Regenerative and Cellular Therapy Products, Gene

Therapy Products, and Cosmetics. On the basis of the inspection and assessment, PMDA concluded that there

were no obstacles to conducting its review based on the application documents submitted.

3. Overall Evaluation

As a result of the above review, PMDA has concluded that the product may be approved after modifying the

proposed indications and the proposed dosage regimen as shown below, with the following condition of

approval, provided that the necessary precautionary statements are included in the package insert and

information on the proper use of the product is properly disseminated after the market launch, and provided

that the product is used under the supervision of physicians with sufficient knowledge and experience in cancer

100


chemotherapy at medical institutions capable of emergency response. The re-examination periods for the

present partial change applications are as follows:

• Malignant melanoma (The re-examination period is the reminder of the ongoing re-examination period

[until September 27, 2026].)

• Unresectable, advanced or recurrent non-small cell lung cancer (The re-examination period is the reminder

of the ongoing re-examination period [until October 18, 2022].)

• Advanced or recurrent MSI-High solid tumors that have progressed after cancer chemotherapy (limit the

use to patients who are refractory or intolerant to standard treatments) (The re-examination period is 4

years.)

Indications (Underline denotes additions. Strikethrough denotes deletions.)

Unresectable Malignant melanoma



Radically unresectable urothelial carcinoma that has progressed after cancer chemotherapy

Advanced or recurrent microsatellite instability-high (MSI-High) solid tumors that have progressed after

cancer chemotherapy (limit the use to patients who are refractory or intolerant to standard treatments)

Dosage and administration (Underlines denote additions. Strikethrough denotes deletions.)


The usual adult dosage is 200 mg 2 mg/kg body weight of pembrolizumab (genetical recombination) infused

intravenously over 30 minutes every 3 weeks. For the adjuvant treatment of malignant melanoma, the

maximum duration of treatment is 12 months.

[Unresectable, advanced or recurrent PD-L1 positive non-small cell lung cancer, relapsed or refractory cHL,

radically unresectable urothelial carcinoma that has progressed after cancer chemotherapy, or advanced or

recurrent MSI-High solid tumors that have progressed after cancer chemotherapy (limit the use to patients who

are refractory or intolerant to standard treatments)]

The usual adult dosage is 200 mg of pembrolizumab (genetical recombination) infused intravenously over 30

minutes every 3 weeks.


[Malignant melanoma, unresectable, advanced or recurrent non-small cell lung cancer]





101


2. The applicant is required to provide healthcare professionals with the results of 2 ongoing phase II studies,




tumors (except colorectal cancer), the applicant is required to conduct a use-results survey after the market

launch to collect information on the characteristics of patients treated with the product, to promptly collect

data on the efficacy and safety of the product, and to take necessary actions for the proper use of the

product.

Warning (no change)

(1) Pembrolizumab should be administered only to patients considered appropriate to receive

pembrolizumab therapy by a physician with sufficient knowledge and experience in cancer

chemotherapy at a medical institution that can provide adequate emergency medical care. Inform the

patient or their family members of the effectiveness and risks of pembrolizumab and obtain their

consent before the start of treatment.

(2) There have been reports of patients who died after experiencing interstitial lung disease. Patients

should be closely monitored for initial symptoms (shortness of breath, dyspnoea, cough, etc.) and

examined by chest X-rays. If any abnormalities are observed, discontinue pembrolizumab and take

appropriate measures such as treatment with corticosteroids.

Contraindications (no change)

Patients with a history of hypersensitivity to any ingredient of the product

Precautions for Indications (Underlines denote additions. Strikethrough denotes deletions.)


The efficacy and safety of pembrolizumab in adjuvant chemotherapy have not been established.

Appropriate patients should be selected based on a good understanding of the efficacy and safety of


characteristics (e.g., disease stage) of patients enrolled in the clinical studies.

[Unresectable, advanced or recurrent PD-L1 positive non-small cell lung cancer]

(1) The efficacy and safety of pembrolizumab in adjuvant chemotherapy have not been established.

(2) When used as monotherapy, pembrolizumab should be administered to patients whose tumors have

been demonstrated to express PD-L1. The “Clinical Studies” section should be carefully read to

understand the percentage of PD-L1-expressing tumor cells (TPS). Pembrolizumab should be

administered to patients whose tumors have been demonstrated to express PD-L1. PD-L1 testing

should be performed by a highly experienced pathologist or at a laboratory facility using the approved

in vitro diagnostic.

102


(3) Appropriate patients should be selected based on a good understanding of the efficacy and safety of

pembrolizumab, after carefully reading the “Clinical Studies” section to understand the status of

EGFR mutation or ALK fusion genes in patients enrolled in the clinical studies.

[Relapsed or refractory cHL]

Appropriate patients should be selected based on a good understanding of the efficacy and safety of


characteristics (e.g., prior treatments) of patients enrolled in the clinical studies.

[Radically unresectable urothelial carcinoma that has progressed after cancer chemotherapy]

(1) The efficacy and safety of pembrolizumab as the first-line therapy have not been established.


pembrolizumab, after carefully reading the “Clinical Studies” section to understand particularly for

the characteristics (e.g., prior treatments) of patients enrolled in the clinical studies.

(3) The efficacy and safety of pembrolizumab in adjuvant chemotherapy have not been established.



(1) Pembrolizumab should be administered to patients who have been demonstrated to have an advanced

or recurrent MSI-High solid tumor by a highly experienced pathologist or at a laboratory facility

using the approved in vitro diagnostic.

(2) In the treatment of colorectal cancer, the efficacy and safety of pembrolizumab have not been

established in patients with no prior therapy with a fluoropyrimidine + oxaliplatin + irinotecan

hydrochloride hydrate.

(3) In the treatment of solid tumors (except colorectal cancer), the efficacy and safety of pembrolizumab

as the first-line therapy have not been established. In the second-line therapy for these solid tumors,

standard treatments, if available, should be chosen before pembrolizumab.

(4) The efficacy and safety of pembrolizumab in adjuvant therapy have not been established.


pembrolizumab, after carefully reading the Clinical Studies” section to understand particularly for

the characteristics (e.g., cancer type) of patients enrolled in the clinical studies as well as carefully

considering the choice of alternative therapies.

Precautions for Indications (Underlines denote additions.)

[Unresectable, advanced or recurrent non-small cell lung cancer]

When pembrolizumab is used with other antineoplastic drugs, the “Clinical Studies” section should be


103


[Malignant melanoma, relapsed or refractory cHL, radically unresectable urothelial carcinoma that has

progressed after cancer chemotherapy, or advanced or recurrent MSI-High solid tumors that have progressed

after cancer chemotherapy (limit the use to patients who are refractory or intolerant to standard treatments)]

The efficacy and safety of pembrolizumab in combination with other antineoplastic drugs have not been

established.

[All the indications]

If an adverse drug reaction associated with pembrolizumab occurs, interrupt or discontinue pembrolizumab in

accordance with the table below.

Adverse reaction Severity Actions

Interstitial lung disease

Grade 2

Interrupt pembrolizumab until the reaction resolves to Grade ≤1. If the reaction has resolved to Grade ≤1 over a >4-week period, resume pembrolizumab every 3 weeks. Discontinue pembrolizumab if the reaction has not resolved to Grade ≤1 after >12 weeks of interruption.

Grade ≥3 or relapsed Grade 2 Discontinue pembrolizumab.

Colitis or diarrhoea

Grade 2 or 3

Interrupt pembrolizumab until the reaction resolves to Grade ≤1. Discontinue pembrolizumab if the reaction has not resolved to Grade ≤1 after >12 weeks of interruption.

Grade 4 Discontinue pembrolizumab.

Hepatic function disorder

AST (GOT) or ALT (GPT) increased to 3-5 times the upper limit of normal (ULN), or total bilirubin increased to 1.5-3 times the ULN

Interrupt pembrolizumab until the value decreases to below the criteria on the left. Discontinue pembrolizumab if the value has not decreased to below the criteria on the left after >12 weeks of interruption.

• AST (GOT) or ALT (GPT) increased to >5 times the ULN, or total bilirubin increased to >3 times the ULN

• Patients with liver metastasis: Grade 2 AST (GOT) or ALT (GPT) at baseline with a ≥50% increase from baseline persisting for ≥1 week

Discontinue pembrolizumab.

Renal impairment Grade 2

Interrupt pembrolizumab until the reaction resolves to Grade ≤1. Discontinue pembrolizumab if the reaction has not resolved to Grade ≤1 after >12 weeks of interruption.

Grade ≥3 Discontinue pembrolizumab.

Endocrine disorders

• Grade ≥2 hypophysitis • Symptomatic endocrine disorders (except for

hypothyroidism) • Grade ≥3 thyroid dysfunction • Grade ≥3 hyperglycaemia • Type 1 diabetes mellitus

Interrupt pembrolizumab until the reaction resolves to Grade ≤1. Consider discontinuing pembrolizumab if the reaction has not resolved to Grade ≤1 after >12 weeks of interruption.

Infusion reaction Grade 2

Stop infusion immediately. If the reaction has resolved within 1 hour, resume pembrolizumab by reducing the infusion rate by 50%.

Grade ≥3 or relapsed Grade 2 Stop infusion immediately, and do not resume pembrolizumab.

Other adverse reactions

• Grade 4 adverse reactions • If the dose of corticosteroid to treat an adverse

reaction cannot be reduced to the dose equivalent to ≤10 mg/day prednisolone within 12 weeks of start of treatment:

• Adverse reactions that have not resolved to Grade ≤1 after >12 weeks of interruption

Discontinue pembrolizumab. If Grade 4 hematotoxicity occurs in a patient with relapsed or refractory cHL, interrupt pembrolizumab until the reaction resolves to Grade ≤1.

Grade is determined in accordance with the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE) v4.0.

i


Appendix


a platinum CBDCA or CDDP a platinum + PEM a combination of a platinum-based chemotherapy and PEM ALK anaplastic lymphoma kinase ALP alkaline phosphatase ALT alanine aminotransferase AST aspartate aminotransferase AUCss,6wk

area under the concentration-time curve at steady state over a 6-week interval

AUC5 area under the concentration-time curve over a 5-week interval AUC6 area under the concentration-time curve over a 6-week interval bevacizumab bevacizumab (genetical recombination) CBDCA carboplatin CBDCA/nab-PTX a combination of CBDCA nab-PTX CBDCA/PEM a combination of CBDCA and PEM CBDCA/PTX a combination of CBDCA and PTX CDDP cisplatin CDDP/PEM a combination of CDDP and PEM cetuximab cetuximab (genetical recombination) cHL classical Hodgkin lymphoma CI confidence interval CPS combined positive score CPT-11 irinotecan hydrochloride hydrate CR complete response dabrafenib dabrafenib mesilate dabrafenib + trametinib a combination of dabrafenib and trametinib DLT dose limiting toxicity DMFS distant metastasis-free survival DOC docetaxel hydrate dMMR deficient mismatch repair EGFR epidermal growth factor receptor FOLFIRI a combination of 5-FU, calcium folinate (or levofolinate), and

CPT-11 FOLFOX a combination of 5-FU, calcium folinate (or levofolinate), and L-

OHP FOLFOXIRI a combination of 5-FU, calcium folinate (or levofolinate), L-OHP,

and CPT-11 GGT gamma-glutamyltransferase IHC immunohistochemistry ILD interstitial lung disease ipilimumab ipilimumab (genetical recombination) IRR infusion related reaction ITT intention-to-treat Japanese clinical practice guidelines (large intestine cancer)

Japanese Society for Cancer of the Colon and Rectum (JSCCR) Guidelines 2016 for the Treatment of Colorectal Cancer

Japanese clinical practice guidelines (malignant melanoma)

Scientific Evidence-based Cutaneous Malignant Tumor Treatment Guidelines, Version 2. edited by the Japanese Skin Cancer Society

Japanese clinical practice guidelines (NSCLC)

Guidelines for Diagnosis and Treatment of the Lung Cancer 2017, edited by the Japan Lung Cancer Society

KRAS Kirsten rat sarcoma viral oncogene homolog

ii


L-OHP oxaliplatin MedDRA MedicalDictionary for Regulatory Activities MEL score ≤1 melanoma score ≤1 MEL score ≥2 melanoma score ≥2 MLH1 mutL homolog 1 MSH2 mutS homolog 2 MSH6 mutS homolog 6 MSI microsatellite instability MSI-High microsatellite instability-high nab-PTX nanoparticle albumin-bound paclitaxel NCCN guidelines (Bone Cancer) National Comprehensive Cancer Network Clinical Practice

Guidelines in Oncology, Bone Cancer NCCN guidelines (Cervical Cancer) National Comprehensive Cancer Network Clinical Practice

Guidelines in Oncology, Cervical Cancer NCCN guidelines (Colon Cancer) National Comprehensive Cancer Network Clinical Practice

Guidelines in Oncology, Colon Cancer NCCN guidelines (Esophageal and Esophagogastric Junction Cancers)

National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Esophageal and Esophagogastric Junction Cancers

NCCN guidelines (Gastric Cancer) National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Gastric Cancer

NCCN guidelines (Hepatobiliary Cancer)

National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Hepatobiliary Cancer

NCCN guidelines (Melanoma) National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Melanoma

NCCN guidelines (Non-Small Cell Lung Cancer)

National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Non-Small Cell Lung Cancer

NCCN guidelines (Ovarian Cancer) National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Ovarian Cancer

NCCN guidelines (Pancreatic Adenocarcinoma)

National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Pancreatic Adenocarcinoma

NCCN guidelines (Prostate Cancer) National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Prostate Cancer

NCCN guidelines (Rectal Cancer) National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Rectal Cancer

NCCN guidelines (Uterine Neoplasms)

National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Uterine Neoplasms

NCI-PDQ National Cancer Institute Physician Data Query NE not evaluable nivolumab nivolumab (genetical recombination) NSCLC non-small cell lung cancer NSQ non-squamous NSQ-NSCLC non-squamous non-small cell lung cancer OS overall survival panitumumab panitumumab (genetical recombination) partial change application application of partial change approval PCR polymerase chain reaction PD progressive disease PD-L programmed cell death-ligand PD-1 programmed cell death-1 PEM pemetrexed sodium hydrate pembrolizumab pembrolizumab (genetical recombination)

iii


pembrolizumab + CBDCA + nab-PTX

a combination of pembrolizumab, CBDCA, and nab-PTX

pembrolizumab + CBDCA + PEM a combination of pembrolizumab, CBDCA, and PEM pembrolizumab + CBDCA + PTX a combination of pembrolizumab, CBDCA and PTX pembrolizumab + PEM a combination of pembrolizumab and PEM pembrolizumab + a platinum + PEM a combination of pembrolizumab, a platinum, and PEM PFS progression free survival PK pharmacokinetics PK/PD pharmacokinetics/pharmacodynamics placebo + PEM a combination of placebo and PEM PMDA Pharmaceuticals and Medical Devices Agency PMS2 postmeiotic segregation increased 2 PPK population pharmacokinetics PR partial response PT preferred term PTX paclitaxel Q2W quaque 2 weeks Q3W quaque 3 weeks RECIST Response Evaluation Criteria in Solid Tumors regorafenib regorafenib hydrate RFS Recurrence free survival RMP risk management plan RNA ribonucleic acid SD stable disease SQ squamous SQ-NSCLC squamous non-small cell lung cancer Study 001 Study KEYNOTE-001 Study 002 Study KEYNOTE-002 Study 006 Study KEYNOTE-006 Study 010 Study KEYNOTE-010 Study 011 Study KEYNOTE-011 Study 012 Study KEYNOTE-012 Study 021 Study KEYNOTE-021 Study 024 Study KEYNOTE-024 Study 025 Study KEYNOTE-025 Study 028 Study KEYNOTE-028 Study 037 Study KEYNOTE-037 Study 041 Study KEYNOTE-041 Study 042 Study KEYNOTE-042 Study 045 Study KEYNOTE-045 Study 052 Study KEYNOTE-052 Study 054 Study KEYNOTE-054 Study 055 Study KEYNOTE-055 Study 059 Study KEYNOTE-059 Study 064 Study KEYNOTE-064 Study 087 Study KEYNOTE-087 Study 158 Study KEYNOTE-158 Study 164 Study KEYNOTE-164 Study 177 Study KEYNOTE-177 Study 189 Study KEYNOTE-189 Study 204 Study KEYNOTE-204 Study 407 Study KEYNOTE-407 TPS tumor proportion score

iv


trametinib trametinib dimethyl ulfoxide 5-FU 5-fluorouracil

report on the deliberation results - 医薬品医療機器総合機構 · 2020-07-20 · market...

Documents